Environmental Simulation Chambers: Application to Atmospheric Chemical Processes (NATO Science Series: IV: Earth and Environmental Sciences)

Environmental Simulation Chambers: Application to Atmospheric Chemical Processes

NATO Science Series A Series presenting the results of scientific meetings supported under the NATO Science Programme. The Series is published by IOS Press, Amsterdam, and Springer in conjunction with the NATO Scientific Affairs Division

Sub-Series I. II. III. IV.

Life and Behavioural Sciences Mathematics, Physics and Chemistry Computer and Systems Science Earth and Environmental Sciences

IOS Press Springer IOS Press Springer

The NATO Science Series continues the series of books published formerly as the NATO ASI Series. The NATO Science Programme offers support for collaboration in civil science between scientists of countries of the Euro-Atlantic Partnership Council. The types of scientific meeting generally supported are “Advanced Study Institutes” and “Advanced Research Workshops”, and the NATO Science Series collects together the results of these meetings. The meetings are co-organized bij scientists from NATO countries and scientists from NATO’s Partner countries –countries of the CIS and Central and Eastern Europe. Advanced Study Institutes are high-level tutorial courses offering in-depth study of latest advances in a field. Advanced Research Workshops are expert meetings aimed at critical assessment of a field, and identification of directions for future action.

As a consequence of the restructuring of the NATO Science Programme in 1999, the NATO Science Series was re-organized to the four sub-series noted above. Please consult the following web sites for information on previous volumes published in the Series. http://www.nato.int/science http://www.Springer.com http://www.iospress.nl http://www.wtv-books.de/nato-pco.htm

Series IV: Earth and Environmental Series – Vol. 62

Environmental Simulation Chambers: Application to Atmospheric Chemical Processes edited by

Ian Barnes Bergische University Wuppertal, Physical Chemistry Department, Germany and

Krzysztof J. Rudzinski Institute of Physical Chemistry of the PAS, Warsaw, Poland

Published in cooperation with NATO Public Diplomacy Division

Proceedings of the NATO Advanced Research Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes Zakopane, Poland 1 - 4 October 2004

A C.I.P. Catalogue record for this book is available from the Library of Congress. ISBN-10 ISBN-13 ISBN-10 ISBN-13 ISBN-10 ISBN-13

1-4020-4231-0 (PB) 978-1-4020-4231-7 (PB) 1-4020-4230-2 (HB) 978-1-4020-4230-0 (HB) 1-4020-4232-9 (e-book) 978-1-4020-4232-4 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands.

www.springer.com

Printed on acid-free paper

All Rights Reserved © 2006 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work Printed in the Netherlands.

CONTENTS

Preface

ix

Acknowledgements

xiii

Simulation Chamber Facilities Overview on the Development of Chambers for the Study of Atmospheric Chemical Processes Becker, K. H.

1

The UCR EPA Environmental Chamber Carter, W. P. L.

27

Investigations of Secondary Organic Aerosol in the UCR EPA Environmental Chamber Cocker II, D. R. and Song, C.

43

Field Measurement and Modelling Motivated Design of a Programme of Aerosol Chamber Experiments McFiggans, G.

49

Chamber Simulations of Cloud Chemistry: The AIDA Chamber Wagner, R., Bunz, H., Linke, C., Möhler, O., Naumann, K-H., Saathoff, H., Schnaiter, M., and Schurath, U.

67

New on-line Mass Spectrometer for Identification of Reaction Products in the Aqueous Phase: Application to the O H-oxidation of N-methyl-pyrrolidone under Atmospheric Conditions Poulain, L., Monod, A., and Wortham, H.

83

Dynamic Chamber System to Measure Gaseous Compounds Emissions and Atmospheric-Biospheric Interactions Aneja, V. P., Blunden, J., Claiborn, C. S., and Rogers, H. H.

97

Chemical Process Studies Chamber Studies on the Photolysis of Aldehydes Wenger, J. C.

111

Determination of Photolysis Frequencies for Selected Carbonyl Compounds in the EUPHORE Chamber Olariu, R-I., Duncianu, M., Arsene, C., and Wirtz, K.

121

v

vi

Contents

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy (DOAS): Advancements in Simulation Chamber and Field Experiments Volkamer, R., Barnes, I., Platt, U., Molina, L. T., and Molina, M. J.

129

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies Bloss, C., Jenkin, M. E., Bloss, W. J., Rickard, A. R., and Pilling, M. J.

143

FT-IR Kinetic Study on the Gas-Phase Reactions of the OH Radical with a Series of Nitroaromatic Compounds Bejan, I., Barnes, I., Olariu, R., Becker, K. H. and Mocanu, R.

155

Atmospheric Fate of Unsaturated Ethers Mellouki, A.

163

Atmospheric Oxidation of the Chlorinated Solvents, 1,1,1-Trichloroethane, Trichloroethene and Tetrachloroethene Nolan, L., Guihur, A-L., Manning, M., and Sidebottom, H.

171

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals: Detection of Reaction Products by CW-CRDS in the Near-Infrared Range Crunaire, S., Fittschen, C., Lemoine, B., Tomas, A., and Coddeville, P.

181

Kinetics, Products and Mechanism of O(3P) Atom Reactions with Alkyl Iodides Barnes, I.

193

Kinetics of the Reaction between CF2O and CH3OH Burgos Paci, M. A. and Argüello, G. A.

207

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System Chiappero, M. S., Malanca, F. E., Argüello, G. A., Nishida, S., Takahashi, K., Matsumi, Y., Hurley, M. D., and Wallington, T. J.

213

Kinetic Study of the Temperature Dependence of the OH Initiated Oxidation of Dimethyl Sulphide Albu, M., Barnes, I., and Mocanu, R.

223

Environmental Chamber Studies of Ozone Formation Potentials of Volatile Organic Compounds Carter, W. P. L.

231

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3, Using Environmental Chamber Data: Butane and Its Degradation Products Pinho, P. G., Pio, C. A., and Jenkin, M. E.

241

Studies on Nitrate-Affected SO 2 Oxidation and Their Perspectives Pasiuk-Bronikowska, W. and Bronikowski, T.

253

Contents

vii

Heterogeneous and Aqueous-Phase Transformations of Isoprene Rudzinski, K. J.

261

Investigation of Atmospheric Transformation of Diesel Emissions in the European Photoreactor (EUPHORE) Zielinska, B., Sagebiel, J., Stockwell, W., McDonald, J., Seagrave, J-C., Wiesen, P., and Wirtz, K.

279

Investigation of Real Car Exhaust in Environmental Simulation Chambers: Results from the INFORMATEX and DIFUSO Projects Wiesen, P.

285

The EUROCHAMP Integrated Infrastructure Initiative Wiesen, P.

295

Air Pollution Studies Survey on Atmospheric Chemistry Research in Some New EU Member States and Candidate Countries Batchvarova, E., Spassova, T., Valkov, N., and Iordanova, L.

301

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany: Input Data for Chemical Mass Balance Modelling Niedojadlo, A., Becker, K. H., Kurtenbach, R., and Wiesen, P.

341

Surface and Total Ozone Over Bulgaria Kolev, S. and Grigorieva, V.

351

Heavy Metals Pollution: An Everlasting Problem Mocanu, R., Cucu-Man, S., and Steinnes, E.

359

Atmospheric Wet Deposition Monitoring in Iasi, Romania Arsene, C., Mihalopoulos, N., Olariu, R-I., and Duncianu, M.

369

Problems of Air Quality in Tashkent City Tolkacheva, G. A.

379

Precipitation Quality in Different Zones of the Tashkent Region in Relation to Photo-Chemical Reactions Smirnova, T. and Tolkacheva, G.

393

Influence of Atmospheric Aerosol Contamination on the Regional Climate in Central Asia Chen, B. B. and Lelevkin, V. M.

403

Assessment of Air Pollution and Ecosystem Buffer Capacity in the Industrial Regions of Ukraine Kharytonov, M. M., Anisimova, L. B., Gritsan, N. P. and Babiy, A. P.

415

viii

Contents

Two Neutral Network Methods in Estimation of Air Pollution Time Series Cigizoglu, H. K., Alp, K., and Kömürcü, M.

421

List of Participants

433

Author Index

441

Subject Index

443

PREFACE Atmospheric pollution has many different detrimental impacts on air quality at urban, regional and global scales. Large volume photoreactors (often referred to as smog or simulation chambers) have been used very effectively to investigate and understand many varied aspects of atmospheric chemistry related to air pollution problems. Photochemical smog formation, which was first observed around 1945 in Los Angeles, is now a major environmental problem for all industrialised and densely populated regions of the world. Over the years many different modelling and experimental tools have been developed to analyse and simulate the complex chemical processes associated with tropspheric photooxidant formation. Work in environmental chambers has played a key role in the development of our understanding of the atmospheric chemistry associated with pollution problems on local, regional and global scales. Chamber observations have also been used in connection with environmental policy issues. In general they are used for validation of atmospheric chemical models, studies of chemical reaction mechanisms and as a direct means to test the possible impact of specific chemical compounds on air quality under simulated ambient conditions New large smog chamber installations have been recently developed in the US (Riverside, California), Europe (Jülich, Germany) and Japan, and a large number of smaller scale laboratory chambers are in operation around the world. Over the years there have been numerous new technical developments related to environmental chamber facilities such as the design of the chambers (e.g. outdoor versus indoor chambers), the techniques applied to control the physical and chemical conditions within them and the analytical facilities that are coupled to the chambers. Some chambers are now equipped with technology allowing measurement of highly reactive radicals such as OH HO2 and NO3, that play pivotal roles in atmospheric chemistry. In addition to the technical developments the scientific applications of chambers are being expanded, for example, the current understanding of secondary organic aerosol formation has been developed principally from environmental chamber studies and chambers have been used even more recently to investigate photolysis processes under ”real” atmospheric conditions. Such new scientific applications are often linked to new technical analytical and other technical developments and environmental chambers offer a perfect platform for testing environmental monitors. The many new scientific applications and technical developments of environmental chambers in the last years created a need for an Advanced Research Workshop where scientists who use chambers (experimentalists, analyticlists as well as modellers) could meet to exchange ideas, experiences and results and document advances in the application of chambers to research. With this background a NATO ARW entitled “Environmental Simulation Chambers: Application to Atmospheric Chemical Processes” was organised and held in the Antalowka Hotel (www.polskietatry.pl.) in Zakopane, Poland from Friday 1st October to Monday 4th October 2004. The major goal of the workshop was to review i) the various types of photoreactors currently in use and ii) the considerable diversity of environmental problems to which photoreactors have been used to successfully investigate. The aims were timely since in Europe within the EU 6th Framework programme a project entitled EUROCHAMP was initiated which had the aim of integrating large scale chamber facilities in Europe and also eventually US facilities. Another goal of the meeting was to promote a better awareness of the scientific community, in particular from NATO partner/Mediterranean dialogue countries and the EU ix

x

Preface

countries, of the availability of such facilities and their considerable potential for investigating problems related not just too atmospheric processes but chemical processes in general. It was particularly hoped that the workshop would attract young scientists from NATO partner countries where such facilities are not widely available but with assistance of the expertise from NATO countries basic experimental setups could easily be realized. At the time of the organisation of the workshop the European Union was expanded by 10 New Member States (Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia). Blugaria, Romania and Turkey are at the time of writing EU applicant countries. Presently, chamber technology is mainly located in western European countries and the US, therefore, in order to actively integrate scientists from eastern countries into the workshop programme a special section was organized with presentations from invited key speakers from several of these countries which showcased many of the acute environmental problems within these countries. A highlight of this section was overview of the research within Eastern European Countries by Dr. E. Batchvarova (Bulgaria) which was specially commissioned by Joint Research Center, Ispra, Italy. The workshop created a unique forum for scientists to present the new scientific applications of environmental chambers as well as the technical developments of these installations, including the advantages and drawbacks of the different types of smog chambers that are in use. Scientists from eastern European countries had the opportunity to learn firsthand about chamber techniques and dialogue with chamber users helped to identify potential applications of the techniques to their own environmental problems. This book contains contributions which give a broad overview of the different types of chamber facilities available and their major applications. The contents has been arranged along three thematic lines – presentation of simulation chamber facilities, chemical process studies, mostly utilising the chamber facilities, and regional studies, mostly indicating the needs for chamber use. Restrictions in the size of the workshop meant that not all the major chambers users could be invited to the meeting but we hope that sufficient reference has been made within the various contributions to these facilities that will allow easy referencing by the reader. (i) The section on Simulation Chambers Facilities begins with an historical overview of chamber development in USA and Europe, discussion of the advantages and disadvantages of using the chambers, the experimental techniques applied and gives exemplary results obtained from chambers in Bonn (now non-operational), Wuppertal and Valencia [Becker]. Detailed overviews are provided of smog chambers in Riverside [Carter; Cocker, aerosol studies], Bayreuth [Zetsch, investigation on POPs], and Karlsruhe [Wagner]. Two untypical yet interesting microchambers are presented, that are being used for studies of aqueous-phase oxidation of N-methyl pyrrolidone (an industrial solvent) [Monod] and for the measurement of fluxes of gaseous compounds between water and air or earth and air [Aneja]. (ii) The section on Chemical Process Studies carried out in the chambers contains a general discussion on the photolysis of atmospheric trace components [Wenger], as well as studies on the photolysis of small carbonyl compounds [Olariu], the photolysis of nitroaromatic compounds [Bejan] and the photolysis of chloral and dichloroacetyl chloride [Sidebottom]. Further contributions in this section describe oxidation processes investigated in the chambers and include: the OH initiated oxidation of benzene, toluene and p-xylene leading to formation of glyoxal, that was monitored by DOAS both in a chamber and in the

Preface

xi

field [Volkammer], an overview on the oxidation of aromatic hydrocarbons [Pilling], reactions of nitroaromatic compounds with OH that contribute to the formation of aerosols [Bejan], the reaction of acetic acid with OH radicals studied with the Continuous-Wave Cavity Ring Down Spectroscopy [Crunaire], and the oxidation of alkyl iodides by O(3P) atoms [Barnes]. Other contributions discuss VOC reactivity and ozone formation potential [Carter], the atmospheric fate of unsaturated ethers [Mellouki], reactions of oxygenated fluorocarbons with NOx and methanol [Arguello; Arguello], evaluation of the MCM v.3 against chamber data [Pinho], and investigations on vehicle ehaust emissions [Wiesen; Zielinska]. Two contributions report on aqueous and heterogeneous chemistry, namely the influence of nitrate ions on the mechanism of S(IV) autoxidation [Pasiuk-Bronikowska], and the heterogeneous and aqueous-phase reactions of isoprene [Rudzinski]. (iii) The section on Regional Studies refers to the geographical area stretching from Central and Eastern Europe to Southern Europe and to Central Asia. The contributing countries were Bulgaria, Germany, Uzbekistan, Kyrghistan, Romania, Turkey and Ukraine. The section starts with contribution which surveys the state (at the time of writing) of atmospheric chemistry research in the new EU member and candidate countries [Batchvarova]. The section covers several important topics. Air quality and pollution in cities is discussed in terms of the contribution of vehicle traffic and solvent use [Niedojadlo], the monitoring and analysis of wet deposition (pH and insoluble fraction) [Arsene], the influence of ozone, dust, nitrogen oxides, carbon oxide and sulphur dioxide, as well as their transformation [Smirnova; Tolkacheva], and the contamination of aerosols [Lelevkin]. Behaviour and effects of heavy metals in the atmosphere are discussed [Mocanu] as well as their impact on the ecosystems [Kharytonov]. The latter contribution also analyses the capacity of ecosystems to cope with air pollution. Other contributions discuss monitoring of ozone and collecting of reference data during a solar eclipse [Kolev], using neural network approach to properly fill the gaps in monitoring data [Cigizoglu], as well as lidar analysis of long range transport of aerosols and source identification [Lelevkin]. In closing, environmental simulation chambers continue to serve as indispensable tools in atmospheric chemistry research. Laboratory and modelling studies are being more frequently coupled to simulation chamber experiments for verification of ideas and for obtaining a deeper insight and understanding of chemical processes. Noticeably, present uses of chambers are not being restricted to the study of gas phase processes, the need for the study of aqueous-phase and heterogeneous studies within a common chamber forum has been recognised and methods are emerging for the study of multiphase systems. Environmental simulation chambers are located mainly in the USA and Western Europe. It would be good if the existing chambers could be made available to researchers outside these confines possibly through common projects. Possible initiatives to improve the global disparity in distribution with newly built chambers is also an alternative which could be explored. Some support for such actions exist within European projects such as ACCENT, EUROCHAMP and INTROP (ESF), which were briefly introduced at the workshop.

Ian Barnes and Krzysztof J. Rudzinski, June 2005

ACKNOWLEDGEMENTS The NATO Science Committee is very gratefully acknowledged for having faith in the venture and providing the initial basis financial support for the workshop. The additional support which was received from other sponsors is also very gratefully acknowledged. Without the support of all the sponsors it would not have been possible to invite and support the 55 participants at the meeting and make the workshop such an all-round success. Financial support of the workshop was by the following organizations/industries: x NATO Scientific and Environmental Affairs Division, Brussels x Institute for Environmental and Sustainability, JRC x European Science Foundation Programme INTROP x BMW - Bavarian Motor Company Munich

xiii

Overview on the Development of Chambers for the Study of Atmospheric Chemical Processes Karl H. Becker Bergische University Wuppertal, FB C, Physical Chemistry, Gauss Str. 20, 42119 Wuppertal, Germany Key Words: Atmospheric chemistry, Hydroxyl radicals, Ozone, Photoreactors, Photosmog

Introduction Photoreactors have been used for quite some time by scientists to study organic photochemistry (Bayes, Blacet, Calvert, Gunning, Hammond, Heicklen, Leighton, Okabe, Noyes, Steacy, and others mentioned in the text book by Calvert and Pitts, 1966). In industry, the photolytic initiation of product synthesis has been known since the beginning of the last century. A large chamber, known as the “Große Bonner Kugel” (Groth et al., 1972), was built in Bonn between 1966 and 1968; its conception was initiated by Groth and Harteck to study air glow reactions at the low pressures prevailing in the upper atmosphere. Afterwards, the chamber was also used to study reactions of tropospheric interest, but the knowledge of tropospheric chemistry at that time was very basic. Photolytic radical sources could not be applied in this reactor, and the radicals had to be introduced from the exterior using discharge flow techniques. Details of the chamber in Bonn will be discussed later in the article. Shortly after the Second World War, the Los Angeles photosmog became a problem which was investigated using relatively large chambers for the simulation of plant damage and health effects, e.g. eye irritation (Hagen-Smit, 1952). In the mid 70-ties, the Pitts group in Riverside (Finlayson-Pitts and Pitts, 1986; 2000) started to build an indoor chamber with the objective of exploring the processes which lead to photosmog formation, see Figure 1. Advances were slow because appropriate analytical techniques still had to be developed. Following this activity, Akimoto duplicated the chamber in Japan (Akimoto et al., 1979), while Atkinson in the Pitts group started to develop very successfully the procedures to measure OH, O3 and NO3 reaction rate constants, after the importance of OH radical reactions for tropospheric chemistry had been accepted. At the same time, others started to use Teflon bags for studying smog reactions under irradiation by natural sunlight, but their results were very much limited to the conditions prevailing in L.A.. Rather important work began at Ford Motor Company where in the late 60-ties Weinstock (1969) promoted the importance of OH reactions in the troposphere. In this laboratory Niki used a rather small photoreactor to develop the application of FTIR spectroscopy for quantitative investigation of atmospheric reactions in the laboratory (Niki et al., 1972; Wu et al., 1976; Niki et al., 1981). IR absorption spectroscopy was applied quite early to study chemical reactions of atmospheric interest (Stephens, 1958; Hanst, 1971), based mainly on mirror systems which allowed long path light absorption (White, 1942; 1976; Herriott et al., 1964, 1965). However, real progress and success came with the application of modern FTIR spectrometers by Niki et al. (1981) and further work in the Pitts’ group to quantitatively measure rate constants and products in photoreactors by long path FTIR absorption. 1 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 1–26. © 2006 Springer. Printed in the Netherlands.

2

Figure 1.

K . H. Becker

Indoor chamber from Riverside, already employing long path FTIR absorption spectroscopy (Source: Finlayson-Pitts and Pitts, 2000).

In the late 60-ties the understanding of atmospheric reactions took a large step forward when OH radicals were identified as the most important oxidizing agent in the troposphere. The discovery resulted in particular from the analysis of the CO budget (Heicklen et al., 1969; Weinstock, 1969, Stedman et al., 1970; Levy, 1971), though the oxidation of CO by OH had already been measured two years earlier (Greiner, 1967). The built-up of a radical chain was understood 10 years later when the rate constant of the fast reaction HO2 + NO o OH + NO2 was measured by several groups (Howard and Evenson, 1977; Leu, 1979; Howard, 1979; Glaschick-Schimpf et al., 1979; Hack et al., 1980; Howard, 1980; Thrush and Wilkinson, 1981) after Crutzen and Howard (1978) indicated the important role of this reaction in stratospheric ozone chemistry. However, it turned out that OH/HO2 radicals are also the key species in the tropospheric oxidation processes that determine the smog formation. In Europe, studies of tropospheric reactions were started in the 70-ties by a few laboratories, e.g. Becker and coworkers in Bonn and Cox and coworkers in Harwell. The technique of long path FTIR absorption mentioned above was introduced in the early 80-ties by Becker and co-workers in Wuppertal. They used a multiple reflection mirror system in a 420 l photoreactor which can be operated between -50 to 50°C to determine the OH reaction rate constants in combination with product analyses in the ppm range, Somewhat later a 6 m long quartz glass reactor of 1080 l volume was built which enabled measurements to be extended down to the ppbV level. The reaction chambers are illustrated in Figures 12 - 16 later in the text where some examples of selected experimental work carried out in the chambers are also given. More recently, other European laboratories have started to use indoor chambers up to 1000 l volume irradiated by black lamps (Baltensperger et al. in Villigen/Zürich, Carlier and Doussin in Paris, Hjorth et al. in Ispra, Herrmann et al. in Leipzig, Le Bras et al. in Orléans, Treacy et al. in Dublin, Wenger et al. in Cork). Tables 1 and 2 show a, not necessarily complete, list of larger indoor and outdoor reactors.

Overview on the Development of Chambers Tabl e 1.

Indoor chambers without light sources, or irradiated by black lamps or solar simulators.

Year of Laboratories Building 1968 1976 1976

3

Groth et al., Bonn Pitts et al., Riverside Niki et al., Dearborn

Description

Application

dark chamber 220 m3, stainless steel (high vacuum) 6 m3, evacuable, thermostated, FEP coated aluminium 150 l Pyrex

without light sources, for low pressure studies photooxidant and kinetic studies kinetic and mechanistic studies

1979

Akimoto et al., 6 m3 evacuable, thermostated, FEP coated aluminium Tsukuba

photooxidant studies

1980

Winer et al., Los Angeles

6 m3 evacuable, thermostated, FEP coated aluminium

photooxidant studies

1981

Becker et al., Wuppertal

420 l Duran glass, evacuable, thermostated -50 to +50 ºC

gas phase studies

1982

Joshi et al. (EPA), Research Triangle Park

440 glass reactor

photooxidant studies

1986

Becker et al., Wuppertal

quartz glass 1080 l, evacuable, thermostated 0 to +25 °C

1986

Evans et al., Australia

4 x 200 l FEP bags

gas phase and aerosol kinetic and mechanistic studies photooxidants studies

1988

Zetzsch et al., ca. 3000 l, Duran glass, Hannover, now thermostated –25 oC to ambient temperature Bayreuth

aerosol studies

1990

Schurath et al., 84 m3, thermostated -90 to +60 ºC, AIDA Karlsruhe

without light source, for aerosol studies

1996

Wahner et al., Jülich.

256 m3, FEP wall cover of a lab room

without light source, for NOY chemistry

1997

Carlier et al., Paris

977 l, glass

gas phase mechanistic studies

1998

Seinfeld et al., Pasadena Carter et al., Riverside

2 × 28 m3, 10 to 40 ºC

aerosol studies

FEP, double wall

low NOx studies

2000

K. H. Becker

4 Table 2.

Outdoor chambers irradiated by sunlight.

Year of Building

Laboratories

Volume, Wall Material

1976

Jeffries et al., Chapel Hill 25 m3, FEP

1981

Lonneman et al.,

15-40 l, FEP bags

1981

Fitz et al.,

40 m3, FEP

1983

Spicer et al.,

17.3 m3, FEP

1985

Kelly et al.,

450-2000 l, FEP bags

1985

Kamens et al., Chapel Hill

25 m3, FEP

1985

Seinfeld et al., Pasadena

65 m3, FEP

1995

Becker et al., Valencia

2 x 200 m3, FEP, EUPHORE

2000

Wahner et al., Jülich

280 m3, FEP, double wall, SAPHIR

In the meantime several large outdoor chambers have been built in the US with support from EPA to study the mechanisms of smog chemistry. One aim was to determine the ozone isopleths under US conditions. These chambers were mainly constructed from bags made of FEP, with volumes up to 25 m3. They brought about a better understanding of smog chemistry, but no results were obtained that could be generalized, because of narrow simulation conditions requested by the US-EPA. In Riverside, Carter and coworkers http://pah.cert.ucr.edu/~carter/bycarter.htm) developed a method to define the ozone ( formation potential of VOC's by determining maximum incremental reactivity (MIR) factors using smog chamber data and chemical modelling. A similar method was introduced by Jeffries in Chapel Hill (http://airsite.unc.edu/), who also used a smog chamber. Other groups injected engine exhausts directly into a smog chambers, and studied the formation of ozone. Collected at very different VOC/NOX ratios, the results from the US could not be generalized and applied to conditions in other countries. In addition, Atkinson and coworkers (http://www.ph.ucla.edu/scpcs/faculty/atkinson.html) refined their method to determine the OH reactivity from chamber measurements, using structure activity relationships to calculate rate constants of OH radical reactions with VOC’s. In the mid 80-ties, Seinfeld (http://www.che.caltech.edu/faculty/seinfeld_j/) in Pasadena, started to use a 65 m3 outdoor chamber made of FEP to study the aerosol formation from the oxidation of aromatic and biogenic hydrocarbons. In the mid 90-ties Becker and co-workers, together with Millán, built in Valencia/Spain the first European large outdoor chamber called EUPHORE (European Photoreactor). Actually, it consists of two chambers made of FEP, each of 200 m3 volume (Becker, 1996). This facility became a centre for European laboratories to carry out mechanistic and kinetic studies. Real exhaust gases from gasoline and Diesel engines were also investigated with respect to ozone formation. The EUPHORE chambers are equipped with all the necessary analytical instrumentation, also for in situ HO2 and OH measurements. In 2000, the group of Ehhalt in Jülich, now guided by Wahner, built a new double wall

Overview on the Development of Chambers

5

outdoor chamber called SAPHIR, of 280 m3 volume, see Figure 2. The double wall made of FEP allows studies of oxidation processes at low NOx concentrations (below 1 ppbV). The Ehhalt group did pioneering work in field measurements of OH and HO2 concentrations, so SAPHIR is fully equipped with the most advanced in situ radical measurement techniques. A smaller double wall indoor chamber (see book contribution) was recently built by Carter in Riverside, to study tropospheric oxidation processes at low NOX concentrations.

Figure 2 .

The double wall outdoor chamber SAPHIR in Jülich, Germany.

Two other chambers were built in Germany, at the same time, for the study of aerosol processes. Already in 1986, Zetzsch and coworkers built in Hannover a 3000 l Duran-glass indoor chamber, covered inside with FEP, and irradiated by solar simulators. This facility has recently been moved to Bayreuth. In 1990, Schurath and co-workers (see contribution in this book) built in Karlsruhe a 84 m3 stainless steel chamber called AIDA, which could be operated between -90 and +60 ºC. EUPHORE will be described later in more detail. Other groups also now operate medium sized chambers: Baltensperger in Zürich, Wenger in Cork, Le Bras in Orleans, Herrmann in Leipzig, Hjorth in Ispra and Doussin in Paris.

6

K. H. Becker

Large static reactors which can be operated at atmospheric pressure Advantages  The probability of wall effects relative to gas reactions decreases with the size R of a reactor, K ~ c • 1/R.  The large reactor diameter favours the application of long path absorption spectroscopy in the UV, VIS and IR. *  Sampling of gas is not limited, and allows the use of a large variety of analytical instruments. *  The large volume inside a reactor provides space for in situ analytical equipment,* e.g. cryo-trapping of radicals and LIF.  Application of the relative rate method enables the determination of rate constants not influenced by impurities, which is a problem in absolute rate determination.  The study of radical reactions with a reactant using the relative rate method is not influenced by radical chains which may develop during the reaction process.  A mixture of reactants can be introduced together with air or inert gases as tracers.  Studies of aerosol formation, and the chemical composition and physical properties of the particles become possible, provided the temperature gradient in the chamber can be controlled. *) This allows the analysis of reaction mechanisms by measuring intermediate products and operation in the ppbV range, i.e. at nearly ambient concentrations.

Disadvantages  With increasing R the volume of the reactor increases by R3 which creates room problems and increases the costs.  Low rate constants for radical reactions, below 10-14 cm3/(molecule • s), cannot be measured because the change of the reactant concentration is given by k×[radical] and [radical] is low, e. g. for OH in the order of 107 -108 radicals/cm3.  The photolytic radical source in a chamber might be less controllable (e.g. wall source of HONO). Special Problems  The light source of indoor chambers may deviate from the sunlight and create different photochemistry.  In the outdoor chamber, the solar spectrum varies with cloud cover and solar zenith angle and the distribution of the radiation is difficult to model. Radical Sources The radical source is a crucial point in using reaction chambers for photooxidation studies. At the successful dawn of chamber work, direct radical measurements were not technically possible. Production of OH radicals in systems like NOX + VOC + air + light occurred mainly by uncontrolled surface processes producing HONO which is photolysed.

Overview on the Development of Chambers

7

Photolytical OH sources in indoor chambers: x

VOC + NOX + air + hQ (source not controlled, probably includes HONO photolysis)

x

HONO + hQ o OH + NO

x

O3 + H2O + hQ o 2 OH + O2

x

RONO + NO + air + hQ o ....... OH RONO + hQ o RO + NO o R'CHO + HO2 RO + O2 HO2 + NO o OH + NO2

x

H2O2 + hQ o 2 OH

x

HCHO + NO + air + hQ o ..... 2 OH + air HCHO + hQ o H + HCO H + O2 + M o HO2 + M HCO + O2 o HO2 + CO HO2 + NO o OH + NO2 Dark OH sources used in chambers: x HO2NO2 + NO + air o OH + 2 NO2 HO2NO2 + M o HO2 + NO2 + M HO2 + NO o OH + NO2 x

olefins + O3 o .....OH......

x

N2H4 + O3 o ..... OH ......

NO3 radical source used in chambers: x N2O5 + M o NO3 + NO2 + M

x NO2 + O3 o NO3 + O2 RO2-radical rource used in chambers: x Cl2 + RH + O2 + hQ + M o 2 RO2 + 2 HCl + M The above lists shows the reactions which have been used to produce OH radicals under controllable conditions by irradiation with an appropriate light source or in the dark. For the study of NO3 radical reactions, either N2O5 has to be introduced into the chamber or the studies have to be carried in the presence of NO2 + O3. For RO2 radical studies, H atom abstraction by photolytically generated Cl atoms in the presence of O2 was employed. For indoor chambers, a variety of photolysis lamps have been used, either intensive Xe high pressure arc or fluorescent lamps with different spectral transmittances. The reactor walls for indoor chambers had to be made from Duran/Pyrex glass or quartz glass in order to ensure transmittance of photolytically active light for initiation of the radical photooxidation (see Figure 3) or the lamps had to be installed inside the reactor. For simulation studies the radical source should be as similar as possible to the outdoor situation, however, in reality this is never fully achieved. Figure 4 gives spectral thresholds for the production of photolytic radical sources in the troposphere, and Figure 5 shows the diurnal variation of the production rates of the different OH radical sources based on field measurements in Germany 1998 within the BERLIOZ campaign. This shows that different OH sources even during daylight

K. H. Becker

8

are active in the troposphere near the surface. Apart from simulation studies photoreactors of all sizes have been successfully used for a long time to measure reaction rate constants of radicals by employing the relative rate technique.

Figure 3 .

Absorptions spectra of NO2, HONO and CH3ONO.

Figure 4 .

Spectral thresholds of photolytic radical sources in the troposphere and the spectral distribution of sunlight a) above the ozone layer, b) near the surface.

Overview on the Development of Chambers

9

Figure 5 . Comparison of the four dominant atmospheric OH radical sources (photolysis of O3, HONO, HCHO, and ozonolysis of VOC’s) for 20th and 21st of July 1998. The difference in the OH production in the early morning hours can be seen easily. After 12:00 on 20th July and 8:00 on 21st July, respectively, the modelled J(HCHO) values were used, because no measurements were available for this time period. The dotted line shows the calculated P(OH) from nitrous acid photolysis during daytime. The data were obtained during the BERLIOZ campaign at the Pabstthum site, northwest from Berlin (Platt et al., 2002).

Relative rate technique The relative rate technique has a few advantages over other methods, i.e. no absolute concentrations have to be measured, impurities do not generally disturb the measurements and the experiments can be carried out in the presence of several reaction partners (Barnes et al., 1982). Also the initiation of radical chains during the reaction process does not distrurb the measurements. However, the rate constant of the reference reaction, k2, has to be known. The following system illustrates the kinetics of the relative rate technique.

10

K. H. Becker X1 + Y Products, k1 (cm3 s-1) X2 + Y Products, k2 X1 and X2: stable species, Y: radical

Reactions:

d [X1] / dt d ln [X1] / dt d [X2] / dt d ln [X2] / dt

= = = =

- k1 [X1] [Y] - k1 [Y] - k2 [X2] [Y] - k2 [Y]

(1/k1) ³d ln [X1]t = - ³[Y]t dt t

t

(1/k2) ³d ln [X2]t = - ³[Y]t dt t

t

(1/k1) ³d ln [X1]t = (1/k2) ³d ln [X2]t t

t

k1 / k2 = ln ([X1]t / [X1]o) / ln ([X2]t / [X2]o)

Selected experimental results from chamber studies "Große Bonner Kugel" The chamber was built between 1966-68 by W. Groth upon the initiative of P. Harteck. The facility was operated by K. H. Becker, E. Fink, D. Kley and U. Schurath (Groth et al., 1972). Properties of the chamber: volume: 221 m3 surface: 177 m2 inner diameter: 7.5 m heatable: Tmax = 350 ºC with 233 KW power for reaching a vacuum of 10-9 Torr. cooling of baffles between pumps and chamber with liquid hydrogen material: stainless steel, 10 mm wall thickness pumping speed: 240 000 l/s with 8 diffusion pumps lowest pressure: 10-9 Torr Major purpose: Studies of O and N atom reactions of aeronomic interest in the upper atmosphere in the pressure range of 0.1-100 mTorr. Major achievements: New results on the recombination reactions of NO + O + (M) (Becker et al., 1972c; 1973), N + N + (M) (Becker et al., 1969; 1971a; 1972b), reactions of O2(1'g) and O2(16+g) with other trace gases (Becker et al., 1971b; 1972a; Penzhorn et al., 1974), reactions of O3 with olefins (Becker et al., 1974a), detection of an energy transfer reaction from O2(1'g) to the HO2 radical and identification of a low lying excited state of HO2 (Becker et al., 1974b; 1975; 1978a). The studies on NO + O and N + N included the quantification of radiative recombination by inverse predissociation which could be clearly

Overview on the Development of Chambers

11

identified in the low pressure regime. Unfortunately, at that time dark OH radical sources and the importance of OH reactions were not known. Figures 6 - 8 show the facility installed beside the Institute of Physical Chemistry, Bonn University. The facility was dismantled few years ago.

Figure 6 .

The housing of the “Große Bonner Kugel”.

Figure 7. Cross-section through the spherical reaction chamber.

12

K . H. Becker

Figure 8 . The pipe system by which the chamber could be heated up to 300 °C (top), enormous pumping capacity needed to reach the vacuum of 10-9 Torr (middle), and the platform at which the experiments were prepared and carried out (bottom).

Overview on the Development of Chambers

13

Studies on the air glow reaction NO + O + M o NO2 + M + hQ The controversial discussion surrounding this reaction as to the relative contributions of a two-body or three-body mechanism could only be settled by experiments carried out at relatively low total pressure, because nearly all previous studies were performed in flow tubes at about 1 Torr or in flow tubes at lower pressures, with significant wall influences. It was expected that the large reaction chamber offered the best possibility to study the reaction at low pressures. The O atoms were generated by flowing O2 through a discharge before introducing the atoms into the chamber. This could be done from the side or even from the centre of the sphere by means of inserted Teflon tubes. After switching off the discharge, the atoms decayed rapidly by wall reactions. However, the large volume provided about 100 s time during the decay, in which the air glow could be measured during stepwise increase of the total pressure. Figure 9 shows an intensity-time profile of the air glow during the stepwise increase of the total pressure.

Figure 9 . The intensity-time profile of the air glow reaction NO + O + M during the decay of the O atoms and the stepwise increase of the total pressure, M = O2, from 0.1 to 0.8 mTorr (upper part) and from 0.1 to 24.5 mTorr (lower part). As can be seen from Figure 11, below 1 mTorr the air glow rate constant follows a two-body radiative recombination with a rate constant of 4.2 × 10-18 cm3 molecule-1 s-1

14

K . H. Becker

Figure 10. Spectral distribution of the air glow emission at different total pressures (M = O2), absolutely calibrated against the distribution at 1 Torr measured in a flow tube (Fontijn et al., 1964).

Figure 11. The air glow intensity obtained by the integral over the spectral distribution of the chemiluminescence shown in Figure 10, which is plotted as apparent two-body rate constant kM against total pressure (M = O2).

Overview on the Development of Chambers

15

whereas above 1 Torr the “high pressure” limit with an apparent two-body rate constant of 6.4 × 10-17 cm3 molecule-1 s-1 has been reached where three-body recombination by M and quenching by M is balanced. From the slope of the plot of kM vs. total pressure a three-body rate constant of 7 × 10-32 cm6 molecule-1 s-1 was determined. As a consequence, the air glowreaction proceeds by a two-body as well as by a three-body reaction. When the air glow rate constant is compared with the total recombination rate of NO + O + M, which also was measured in the chamber, it can be concluded that the air glow rate constant represents within the error limits the total recombination rate constant. Further details are given in the publications Becker et al. (1972c) and (1973). Experiments from Wuppertal Several photoreactors between 420 to 1080 l volumes are in operation in the Physical Chemistry Department of the Bergische University Wuppertal. A thermostated photoreactor of 420 l volume is shown schematically in Figure 12. The reactor consists of an inner Duran glass tube, o.d 0.6 m and length 1.5 m which is concentrically inserted in a wider Duran glass tube of the same length. On both sides the tubes are closed by aluminium flanges covered with Teflon on the inner side. Between the tubes a cooling fluid circulates, and inside the fluid space black lamps/fluorescent lamps in protective Duran covers are installed. The covers are flushed with dry purified air. Figure 13 shows that the whole reactor including the cooling systems is inserted into a rigid optical bench which is part of a White multiple path mirror system coupled to a FTIR spectrometer. In order to minimise light losses of the IR beam, the White mirrors are installed inside the reactor, but can be adjusted to optimal path length from outside. The flanges contain several ports, such as gas inlet/outlet ports and syringe port for taking GC samples, as well as windows for UV/Visible absorption spectroscopy. The optimal path length of several hundred meters enables quantitative concentration measurements in the

Figure 12 . Schematic view of the 420 l reactor in Wuppertal, operating in the temperature range of ± 50 oC.

K . H. Becker

16

ppmV range, while gases with low IR absorption are measured by GC. The spectroscopic measurements require careful recording of reference spectra. The reactor is evacuable down to 1 mTorr, and is operated mostly at atmospheric pressure, filled either with synthetic air or with rare gases. The temperature can be set between –50 to +50 °C. A first version of a

Figure 13.

The suspended double wall reactor of 420 l volume.

Figure 14 .

View of the thermostated and non-thermostated reactors mounted in series.

Overview on the Development of Chambers

17

photoreactor of the same size was built with the cooling system, and now it is installed in line with the thermostated reactor, Figure 14.

Figure 15 . View of the 6 m long quartz glass reactor of 1080 l volume.

Figure 16 .

Schematic drawing of the 1080 l reactor setup.

K . H. Becker

18

Figure 15 and 16 show the 1100 l quartz glass reactor consisting of two combined tubes with a combined length of 6.2 m length with an o.d. of 0.47 m. The end flanges contain all the necessary ports for gas inlets, gas sampling and the White mirrors. The path length for FTIR absorption was extended to approximately 500 meters, to achieve a sensitivity for product analysis reaching the low ppbV range. Examples of experimental work using the chambers (i)

Measurement of pressure dependence of the reaction rate constant of C2H4 + OH in the presence of air and argon (Klein et al., 1984),

(ii)

Measurement of the temperature and pressure dependence of the reaction rate constant HO2NO2 + OH (Barnes et al., 1981; 1986),

(iii)

Measurement of thermal decomposition rates of peroxynitrates, RO2NO2 (Zabel et al., 1986; Kirchner et al., 1990; Becker et al., 1993 and Zabel, 1995) as a function of temperature.

(i) Measurements of the pressure dependence of the reaction rate constant k1 for C2H4 + OH +M o products with M = air and Ar A dark source of OH radicals from the thermal decomposition of HO2NO2 in the presence of NO was employed, and the relative rate method was used with reference reactions, n-hexane + OH and n-butane + OH reactions, for which no pressure dependence is

Figure 17. Dependence of C2H4 + OH +M on total pressure at 295 K for M = synthetic air (M = Ar not shown). The solid line through the data points is the calculated falloff curve. The error bar represents the average of 2V errors for all individual data points.

Overview on the Development of Chambers

19

known. Figure 17 shows the experimental results and, together with Figure 18, demonstrates the high quality of the measurements.

Figure 18 . Comparison of literature k1 values near room temperature with the calculated falloff curves from this work (solid line M = Ar, broken line M = synthetic air. The data points are presented without error limits to simplify the comparison.

(ii) Studies on the temperature and pressure dependence of the reaction rate constant HO2NO2 + OH. The next example describes the measurements of the reaction rate constant of peroxynitric acid (HO2NO2) with OH radicals as a function of pressure (He) and temperature, details of which can be found in the literature (Barnes et al., 1981; 1986). It is impressive that for a rather complex chemical system no absolute concentration measurements were needed. A relative rate method with reference to the known reaction rate constant kn-butane + OH was used. HO2NO2 was prepared outside the reactor and introduced together with NO, He and the reference substance (n-butane) into the reaction chamber. The OH radicals steadily produced from the thermal decomposition of HO2NO2, caused a decay in n-butane which could be followed by GC analysis, stage I in Figure 19. The decay of HO2NO2 measured by FTIR absorption is due to thermal decomposition, possible wall losses and the HO2NO2 + OH reaction, also shown in Figure 19, stage I. After stage I, the n-butane concentration was rapidly increased, which resulted in a change of the OH concentration because of the faster radical consumption by n-butane. Consequently, the decay of HO2NO2 became slower because of the lower OH concentration. From the different slopes of HO2NO2 and n-butane in stage I and stage II, all measured relatively during the time steps 't in both stages, the

20

K . H. Becker

following equation was derived from which the reaction rate constant kHO2NO2 evaluated:

+ OH

was

kHO2NO2+OH/kn-butane+OH = {('ln[HO2NO2]/'t)stage I  ('ln[HO2NO2]/'t)stage II}/ {('ln[n-butane]/'t)stage I  ('ln[n-butane]/'t)stage II}

Figure 19. Concentration-time profiles for X = HO2NO2 (o) and X = n-butane (); reaction conditions: T = 278 K, ptotal = 133 mbar, M = He, [NO]o = 2.4 x 1015 molecules/cm3 , [NO2 ]o = 1.0 x 1014 molecules/cm3. (iii) The thermal decomposition rates of peroxynitrates, RO2NO2, as a function of temperature and pressure. The last example deals with measurements of the thermal decay of peroxynitrates, RO2NO2, as a function of pressure and temperature. Details can be found in the publications Zabel et al. (1986), Kirchner et al. (1990), Becker et al. (1993) and Zabel (1995). In the experiments, the peroxynitrates were produced in the presence of O2 and NO2 from R radicals generated by H atom abstraction by Cl atoms from different hydrocarbons (see the reaction scheme in Figure 20).

Overview on the Development of Chambers

21

Alkane + Cl2 +O2 + M + hQ

ĺ

RO2 +HCl +M

RO2 + NO

ĺ

RO + NO2 k1

ĺ

RONO2

RO2 + NO2 + M

ĺ

RO2NO2 + M

k2

RO2NO2 + M

ĺ

RO2 + NO2 + M

k3

(+M)

k3(effective) = k3 x {1 + (k2/k1)x[NO2]/[NO]}-1 k3(T, [M]) for [NO] >> [NO2] and (k2/k1) from variation of [NO]

Figure 20. Production and decomposition of peroxynitrates in the presence of Cl2, NO, NO2 and O2 during the photolysis of Cl2.

Figure 21. Decay of RO2NO2, here HO2NO2, at different NO concentrations.

According to the equation in Figure 20, by increasing the concentration of NO, the parameter k3 can be determined as a function of temperature and total pressure. This can be seen in Figure 21 where k3 is derived from the slope of ln(ct/co) a t at higher NO concentration. Table 3 shows the results for different peroxynitrates. It can be seen that a

22

K . H. Becker

carbonyl group beside -O2 stabilizes the peroxynitrate as for peroxyacetylnitrat, CH3C(O)O2NO2, (PAN), in particular at lower temperatures. Table 3 .

Rate constants, Arrhenius parameters and thermal lifetimes of several peroxynitrates (experiments and evaluation are described in Zabel et al. (1986), Kirchner et al. (1990), Becker et al. (1993) and Zabel (1995).

R in RO2NO2 C4H9 C6H13 C8H17 C2H5 CH3C(O)CH2 CH3 C6H5CH2 HOCH2 CH3OCH2 CH3C(O)OCH(CH3) C6H5OCH2 H CH3OC(O) CH3C(O) C6H5C(O) *

k3 (298K)*

Ea

Thermal lifetime, K-1

A

-1

kJ/mol

10 s

25 °C

0 °C

- 20 °C

6.5 5.8 3.7 3.7 2.6 1.65 1.5 1.0 0.44 0.24 0.23 0.083 0.00084 0.00040 0.00031

86.2 86.2 86.2 86.2 85.6 83.8 90.3 85.3 92.2 94.9 95.0 89.2 107.0 112.9 116.4

0.83 0.75 0.48 0.47 0.26 0.08 1.0 0.09 0.64 1 1 0.036 0.48 2.5 7.9

0.15 s 0.17 s 0.27 s 0.27 s 0.38 s 0.61 s 0.67 s 1.00 s 2.27 s 4.17 s 4.35 s 12.05 s 19.8 min 41.7 min 53.8 min

3.7 s 4.1 s 6.5 s 6.6 s 9.2 s 13.5 s 19 s 23.3 s 1.1 min 2.4 min 2.5 min 5.4 min 17.2 h 1.8 days 2.7 days

1.3 min 1.4 min 2.2 min 2.2 min 3.0 min 4.2 min 7.3 min 7.7 min 28.3 min 1.1 h 1.1 h 2.0 h 30 days 94 days 157 days

s

16 -1

k3 (T) = A×exp(-Ea/RT)

EUPHORE: the EUropean PHOtoREactor in Valencia In 1995, at the CEAM Foundation in Valencia, the European photoreactor came into operation. The facility was built by the group of Becker from Wuppertal, and is hosted and operated by CEAM under the direction of Millían and Wirtz. Figures 22 and 23 show photos of the facility, which consists of two half-sphere outdoor chambers made from TEDLAR. Each chamber has a volume of 200 m3. Movable rigid covers protect both chambers against heavy storms, and are used to cut off the solar irradiation during experiments. The chambers are fully equipped with state-of-the-art analytical instrumentation. Besides more conventional methods, FTIR long path absortion and DOAS are employed. In one chamber OH and HO2 radicals can be measured in situ by the LASER fluorescence technique. Regular EUPHORE reports are published which can be obtained from Ian Barnes ([email protected]) or Klaus Wirtz ([email protected]).

Overview on the Development of Chambers

Figure 22 . One of the two EUPHORE chambers at CEAM/Valencia.

Figure 23 . Building hosting two EUPHORE chambers, one closed by a metal cover.

23

24

K . H. Becker

Figure 24 . In situ measurements of OH radicals (line) during the oxidation of toluene, for comparison, the points represent OH concentrations that were calculated from the decay rate of toluene (Pilling, M. J. et al., 2003). An example of the good agreement which has been obtained in the EUPHORE chamber between measured OH radical profiles using the FAGE technique (Pilling, M. J. et al., 2003) and the OH radical concentration in the chemical system calculated from the decay of toluene is shown in Figure 24. Outlook Numerous results from kinetic and mechanistic studies using photo-reactors have been obtained which have contributed significantly to an increase in our knowledge on atmospheric chemistry, in particularly in the gas phase. Some open questions still remain, e.g. the mechanisms of the photooxidation of aromatic hydrocarbons and the process of gas-toparticle conversion. More recently also aerosol studies are being carried out in chambers. Some very recent work has led now to a much better understanding of the photochemical wall effects in chambers which produce OH radicals (Rohrer et al., 2005). The state of chemical knowledge has, however, so much advanced that it is quite difficult to propose further innovative work and the interest on additional applications of new results for improving tropospheric models is decreasing. On the other side, interest in exposure experiments using advanced simulation facilities and analytical techniques to study impacts of air pollutants on human health, plant damages and material destruction is on the increase. This could be viewed perhaps as a return to the 50-ties (Hagen-Smit, 1952), however, compared to then present-day approaches are at a much higher level of understanding of the chemical systems and use techniques which enable in situ radical measurements and the monitoring of trace gases in the pptV range. At the time of writing a new EU infrastructure project EUROCHAMP (co-ordinator P. Wiesen < [email protected] >) has as its main goal the integration of all European chamber facilities to improve studies on atmospheric chemical systems and also to open simulation facilities for impact studies.

Overview on the Development of Chambers

25

References Akimoto, H., M. Hoshino, G. Inoue, F. Sakamaki, N. Washida and M. Okuda; Environ. Sci. Technol. 13 (1979) 471. Barnes, I., V. Bastian, K.H. Becker, E.H. Fink and F. Zabel; Atmos. Environ. 16 (1982) 545. Barnes, I., V. Bastian, K. H. Becker, E. H. Fink and F. Zabel; Chem. Phys. Lett. 83 (1981) 459. Barnes, I., V. Bastian, K.H. Becker, E.H. Fink and F. Zabel; Chem. Phys. Lett. 123 (1986) 28. Becker, K.H., W. Groth and D. Kley; Z. Naturforschung 24a (1969) 1840. Becker, K.H., A. Elzer, W. Groth and D. Kley; Z. Naturforschung 26a (1971a) 929. Becker, K.H., W. Groth and U. Schurath; Chem. Phys. Lett. 8 (1971b) 259. Becker, K.H., W. Groth and U. Schurath; Chem. Phys. Lett. 14 (1972a) 489. Becker, K.H. E. H. Fink, W. Groth, W. Jud and D. Kley; Faraday Disc., Chem. Soc. 53 (1972b) 35. Becker, K.H., W. Groth and D. Thran; Chem. Phys. Lett. 15 (1972c) 215. Becker, K.H., W. Groth and D. Thran; 14th Symp. (International) on Combustion, Pittsburgh (1973) 353. Becker, K.H., U. Schurath and H. Seitz; Int. J. Chem. Kinet. 6 (1974a) 725. Becker, K.H., E. H. Fink, P. Langen and U. Schurath; J. Chem. Phys. 60 (1974b) 4623. Becker, K.H., E. H. Fink, P. Langen and U. Schurath; 15th Symposium (International) on Combustion, The Combustion Institute, Pittsburgh (1975) 961. Becker, K.H., E. H. Fink, A. Leiss and U. Schurath; Chem. Phys. Lett. 54 (1978a) 191. Becker, K.H., J. Löbel und U. Schurath; Staub-Reinhaltung Luft 38 (1978b) 278. Becker, K.H., F. Kirchner and F. Zabel; in: K.H. Becker and H.Niki (eds), The Tropospheric Chemistry of Ozone in Polar Regions. NATO ASI Series, Vol. 17, Springer Verlag, Berlin Heidelberg (1993) 351. Bridier, I., F. Caralp, H. Loirat, R. Lesclaux, B. Veyret, K. H. Becker, A. Reimer and F. Zabel; J. Phys. Chem. 95 (1991) 3594. Calvert, J.G. and J.N. Pitts, Jr.; Photochemistry, John Wiley & Sons, New York (1966). Crutzen, P.J. and C.J. Howard; Pure Appl. Geophys. 116 (1978) 497. Finlayson-Pitts, B.J. and J.N. Pitts, Jr.; Atmospheric chemistry, John Wiley & Sons, New York (1986). Finlayson-Pitts, B.J. and J.N. Pitts, Jr.; Chemistry of the upper and lower atmosphere, Academic Press, New York (2000). Fontijn, A., C.B. Meyer and H.I. Schiff; J. Chem. Phys. 40 (1964) 64. George, C., R. S. Strekowski, J. Kleffmann, K. Stemmler and M. Ammann; Faraday Discuss. 130 (2005) 1-16. Glaschick-Schimpf, I.,A. Leiss, P.B. Monkhouse, U. Schurath, K.H. Becker and E.H. Fink, Chem. Phys. Lett. 67 (1979) 318. Greiner, N. R.; J. Chem. Phys. 46 (1967) 2795. Groth, W., K. H. Becker, G.H. Comsa, A. Elzer, E. Fink, W. Jud, D. Kley, U. Schurath and D. Thran; Naturwissenschaften 59 (1972) 379. Haagen-Smit; Ind. Eng. Chem. 44 (1952) 1342. Hack, W., A. W. Preuss, F. Temps, H.Gg. Wagner and K. Hoyermann; Int. J. Chem. Kinet. 12 (1980) 851. Hanst, P. L.; Adv. Environ. Sci. Technol. 2 (1971) 91. Heicklen, J., K. Westberg and N. Cohen; Report, Center for Air Environmental Studies, (1969) 115-169. Herriott, D., H. Kogelnik and R. Kompfner; Appl. Opt. 3 (1964) 523. Herriott, D., and H.J. Schulte; Appl. Opt. 4 (1965) 883.

26

K . H. Becker

Howard, C.J., and K.M. Evenson; Geophys. Res. Lett. 4 (1977) 437. Howard, C.J.; J. Chem. Phys. 71 (1979) 2352. Howard, C.J.; J. Am. Chem. Soc. 102 (1980) 6937. Klein, Th., I. Barnes, K.H. Becker, E.H. Fink and F. Zabel; J. Phys. Chem. 88 (1984) 5020. Leu, M.-T.; J. Chem. Phys. 70 (1979) 1662. Levy, H., Science 173 (1971) 141. Niki, H., E.E. Daby and B. Weinstock; Adv. Chem. Ser., Washington, 113 (1972) 16. Niki, H., P.D. Maker, C.M. Savage and L. P. Beitenbach; Chem. Phys. Lett. 80 (1981) 499. Penzhorn, R.D., H. Güsten, U. Schurath and K.H. Becker; Environ. Sci. Technol. 8 (1974) 907. Pilling, M. J. et al. (2003). Effects of the oXidation of Aromatic Compounds in the Troposphere (EXACT, EVK2-CT-1999-00053), Final Report., http://www.chem.leeds.ac.uk/exact Platt, U., B. Alicke, R. Dubois, A. Geyer, A. Hofzumahaus, F. Holland, M. Martinez, D. Mihelcic, T. Klüpfel, B. Lohrmann, W. Pätz, D. Perner, F. Rohrer, J. Schäfer and J. Stutz; J. Atmos. Chem. 42 (2002) 359. Rohrer, F., B. Bohn, T. Brauers, D. Brüning, F.-J. Johnen, A. Wagner, J. Kleffmann, Atmos.Chem. Phys. Discus. 4 (2004) 7881. Stedman, D.H., E.D. Morris, Jr., E.E. Daby, H. Niki and B. Weinstock; 160th National Meeting of Am. Chem. Soc., Chicago (1970). Stephens, E.R.; Soc. Appl. Spectrosc. 12 (1958) 80. Thrush, B.A. and J.P.T. Wilkinson; Chem. Phys. Lett. 81 (1981) 1. Weinstock, B.; Science 166 (1969) 224. White, J.U.; J. Opt. Soc. Am. 32 (1942) 285. White, J.U.; J. Opt. Soc. Am. 66 (1976) 411. Wu, C.H., S.M. Japar and H. Niki; J. Environ. Sci. Health, Environ. Sci. Eng. A11 (1976) 191. Zabel, F.; Z. Physi. Chem. 188 (1995) 119.

The UCR EPA Environmental Chamber William P. L. Carter College of Engineering Center for Environmental Research and Technology, University of California, Riverside, California, 92521, USA Key Words: Environmental chambers, Atmospheric chemical mechanisms, Ozone, Volatile organic compounds, Oxides of nitrogen

Abstract The UCR EPA chamber is a new large indoor environmental chamber constructed at the University of California at Riverside (UCR) under United States EPA funding for the purpose of evaluating gas-phase and secondary aerosol mechanisms for ground-level air pollution. The major characteristics of this chamber, the results of its initial characterization for gas-phase mechanism evaluation, and examples of initial gas-phase mechanism evaluation experiments, are described. It is concluded that the chamber has lower or at most comparable background effects than other chambers previously used for mechanism evaluation, and can provide useful mechanism evaluation data at NOx levels as low as 2 ppb. Future research directions to utilize the capabilities of this chamber are discussed. Background Chemical mechanisms are critical components of airshed models used for predictions of secondary pollutants such as ground-level ozone or secondary organic aerosol. Because many of the chemical reactions are incompletely understood, these mechanisms cannot be relied upon to give accurate predictions of impacts on emissions on air quality until they have been shown to give accurate predictions under realistic but controlled conditions. The most reliable way to test this is to compare their predictions against results of well-characterized environmental chamber experiments that simulate the range of conditions in the atmosphere. If a model cannot accurately predict results of such experiments, it cannot be expected to reliably predict effects of proposed control strategies on ambient air quality. For this reason, environmental chambers are essential to developing predictive mechanism for compounds for which basic mechanistic information are insufficient (as is the cased for aromatics), testing approximations and estimates necessary for modelling the reactions of almost all VOCs under simulated atmospheric conditions, and testing entire mechanisms under varied conditions. However, results of environmental chamber experiments are only useful for testing mechanisms if the conditions of the experiments are sufficiently well characterized so that the characterization uncertainties are less than the uncertainties of the mechanisms being tested. As discussed by Dodge (2000), the chamber data base used to develop and evaluate current mechanisms had a number of limitations and data gaps that could affect their accuracy. Uncertainties exist concerning characterization of chamber conditions that could cause compensating errors in the gas-phase mechanism (Carter and Lurmann, 1990, 1991; Jeffries et al, 1992). Most chamber experiments lack measurement data for important species, limiting the level of detail to which the mechanisms can be evaluated, and the types of air quality impact predictions that can be assessed. Furthermore, because of chamber effects and because of inadequate analytical equipment employed, the current environmental chamber data base is not suitable for evaluating chemical mechanisms under the lower NOx conditions found in 27 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 27–41. © 2006 Springer. Printed in the Netherlands.

W. P. L. Carter

28

rural and urban areas with lower pollutant burdens. Because of this, one cannot necessarily be assured that models developed to simulate urban source areas with high NOx conditions will satisfactorily simulate downwind or cleaner environments where NOx is low. To address the need for improved an improved environmental chamber facility to evaluate mechanism for O3 and PM formation, the College of Engineering, Center for Environmental Research and Technology (CE-CERT) at the University of California at Riverside (UCR) has undertaken a program to develop a “Next Generation” environmental chamber facility for chemical mechanism evaluation and VOC reactivity assessment. The objectives are to develop the environmental chamber facility needed for evaluating gas-phase and gas-to-particle atmospheric reaction mechanisms, for determining secondary aerosol yields, and for measuring VOC reaction products and radical and NOx indicator species under more realistic and varied environmental conditions than previously has been possible. This project resulted in the construction of new “UCR EPA” chamber, which became fully operational in early 2003. The design features and characteristics of this chamber, the results of its initial characterization, and completed and ongoing projects in this chamber are summarized briefly below. Facility Description The indoor facility comprises a 20’x20’x40’ thermally insulated enclosure that is continually flushed with purified air at a rate of 1000 L min-1 and is located on the second floor of a laboratory building specifically designed to house it. Located directly under the enclosure on the first floor is an array of gas-phase continuous and semi-continuous gas-phase monitors. Within the enclosure are two ~90 m3 (6.1 m x 3.1 m x 5.5 m, Surface area to volume = 1.35 m-1) 2 mil FEP Teflon® film reactors, a 200 kW Argon arc lamp, a bank of 72 W 4-ft blacklights, along with the light and aerosol instrumentation. A schematic of the enclosure is provided in Figure 1.

Movable top frame allows reactors to collapse under pressure control

2 Banks of Blacklights This volume kept clear to maintain light uniformity

200 KW Arc Light

Dual Teflon Reactors

20 Two air Handlers ft. are located in the corners on each side of the light (not shown).

Mixing System Under floor of reactors Floor Frame 20 ft.

Temperature controlled room flushed with purified air and with reflective material on all inner surfaces

Figure 1.

20 ft.

Access Door

SEMS (PM) Instrument

Gas sample lines to laboratory below

Schematic of the environmental chamber reactors and enclosure.

The UCR EPA Environmental Chamber

29

Enclosure The interior of the thermally insulated 450 m3 enclosure is lined with reflective Everbrite® (Alcoa, PA) aluminium sheeting to maximize the interior light intensity and homogenize the interior light intensity. A positive pressure is maintained between the enclosure and the surrounding room to reduce contamination of the reactor enclosure by the surrounding building air. The enclosure air is well mixed by the large air handlers that draw in air from inlets around the light and force the air through a false ceiling with perforated reflective aluminium sheets. The enclosure is temperature controlled with a ~30 ton air conditioner capable of producing a temperature range of -5 to 80 C. Teflon Reactors The 2 mil (54 Pm) FEP Teflon® reactors are mounted within the enclosure with a rigid bottom frame and a moveable top frame. The floor of the reactor is lined with Teflon film with openings for reactant mixing within and between reactors and 8 ports ranging in size from ¼” to ½” for sample injection and withdrawal. The moveable top frame is raised/lowered with a motorized pulley system enabling the user to expand (during filling) and contract (during an experiment or for flushing) as necessary. The rate of contraction/expansion is set to maintain a differential pressure of 0.02 in H2O between the inside of the reactor and the enclosure. During experiments, the top frames are slowly lowered to maintain positive pressure as the volume decreases due to sampling, leaks, and permeation. The experiment is terminated when the final reactor volume reaches 1/3 of its maximum value (typically about 10 hours). The elevator system coupled with differential pressure measurements allows for repeatable initial chamber volumes and allows for reactants to be injected with greater than 5% precision. The Teflon reactors are built in-house using a PI-G36 Pac Impulse Sealer (San Rafael, CA) heat sealing device for all major seams and a U-frame mount to the reactor ceiling and floor. Pure Air System An Aadco 737 series (Cleves, Ohio) air purification system produces compressed air at rates up to 1500 L min-1. The air is further purified by passing through canisters of Purafil® and heated Carulite 300® followed by a filter pack to remove all particulate. The purified air within the reactor has no detectable non-methane hydrocarbons (<1 ppb), NOx (<15 ppt), no detectable particles (<0.2 particles cm-3), and a dew-point below -40 C. The reactors are cleaned between runs by reducing the reactor volume to less than 5% of its original volume while flushing the reactors with 500 L min-1 purified air. The reactor is then returned to its original volume by filling with purified air. No residual hydrocarbons, NOx, or particles are detected after the cleaning process. Light sources A 200 kW Argon arc lamp with a spectral filter (Vortek co, British Columbia, Canada) is used to irradiate the enclosure and closely simulate the entire UV-Vis ground-level solar spectra. The arc lamp is mounted on the far wall from the reactors at a minimum distance of 20’ to provide uniform lighting within both reactors. Backup lighting is provided by banks of 80 1.22 m 115-W Sylvania 350BL blacklamps (peak intensity at 350 nm) mounted on the same wall of the enclosure. These provide a low-cost and efficient UV irradiation source

W. P. L. Carter

30

Relative Intensity (Normalized to give same NO2 photolysis rate)

within the reactor for experiments where the closer spectral match provided by the Argon arc system is not required. The blacklights are located on the ends of the same wall of the Argon arc lamp. Figure 2 shows the spectra for each light source with a comparison to a representative ground-level outdoor solar spectrum.

0.12

0.8

0.10 0.6

0.08 0.06

0.4 UCR EPA Blacklights Solar Z=0

0.04 0.2

0.02 0.00 0.290

0.300

0.310

0.320

0.0 0.300

0.350

0.400

0.450

0.500

0.550

0.600

Wavelength (nm)

Figure 2 .

Spectrum of the argon arc light source used in the chamber. Blacklight and representative solar spectra, with relative intensities normalized to give the same NO2 photolysis rate.

Interreactor and Intrareactor mixing The two reactors are connected to each other through a series of custom solenoid valves and blowers. The system provides for rapid air exchange prior to the start of an experiment ensuring, when desired, that both reactors have identical concentrations of starting material. Each reactor can be premixed prior to the start of an experiment by Teflon coated fans located within the reactor. Analytical Instrumentation The instrumentation used to monitor gas-phase species in this chamber are described by Carter. (2002). Briefly, ozone, CO, NO, and NOy are monitored using commercially available instruments. Two Tuneable Diode Laser Absorption Spectroscopy (TDLAS) are available for monitoring NO2 HNO3, H2O2 and formaldehyde. TDLAS analysis is described in detail elsewhere (Hastie et al., 1983; Schiff et al., 1994) and is based on measuring single rotational - vibrational lines of the target molecules in the near to mid infrared using laser diodes with very narrow line widths and tunability. Two GC-luminol instruments based on previous work by Gaffney et al. (1998) are used to monitor PAN and higher PAN analogues. It uses luminol detection combined with GC column to separate NO2 from PAN and other species that are detected by luminol to provide a specific analysis for those species. Organic reactants other than formaldehyde are monitored by gas chromatography with FID detection.

The UCR EPA Environmental Chamber

31

Instrumentation was also available to measure aerosol size distributions, from which aerosol number and volume can be calculated. This is described by Cocker (2004), and also by Carter (2002). The use of this chamber for aerosol studies is discussed by Cocker (2004). Characterization Results Light Characterization Argon Arc Light. Although the intensity of the argon arc light can be varied by varying the lamp power, normally it is operated at 80% of maximum. Information about trends in light intensity with time is available from data from the spectral radiometer and PAR radiation instruments (Carter et al., 2002), and from results of NO2 actinometry experiments carried out periodically using the quartz tube method of Zafonte et al. (1977) modified as discussed by Carter et al. (1995a). The results indicated no significant change of light intensity with time. Actinometry experiments with the quartz tube located inside the reactors yielded an NO2 photolysis rate of 0.26 min-1. Blacklights. The light intensity inside the reactors with blacklight irradiation was measured in a series of experiments carried out in April-May of 2003 and again in October of that year, and the averages of the results were 0.19 and 0.18 min-1, respectively. The light intensity measured in the enclosure during blacklight experiments indicated a gradual decreasing trend in light intensity during the experiments that was consistent with the differences between these two measurements. Spectra. The spectrum of the arc light sources were measured using a LiCor LI-1800 spectroradiometer, and are shown on Figure 2. No appreciable change in the light source spectrum was observed in the first 18 months of operation. Characterization of Contamination by Outside Air Contamination of the reactor by leaks and permeation of laboratory air contaminants is minimized by continuously flushing the enclosure that houses the reactors with purified air. NOx and formaldehyde levels in the enclosure before or during irradiations were generally less than 5 ppb and PM concentrations are below the detection limits of our instrumentation. Introduction of contaminants into the reactor is also minimized by use of pressure control to assure that the reactors are always held at slight positive pressures with respect to the enclosure, so leaks are manifested by reduction of the reactor volume rather than dilution of the reactor by enclosure air. The leak rate into the chamber was tested by injecting ~100 ppm of CO into the enclosure and monitoring CO within the reactor. No appreciable CO (below the 50 ppb detection limit) was obtained for this experiment. Chamber Effects Characterization It is critical to characterize the impact of reactor walls on gas-phase reactivity and secondary aerosol formation. Larger volume reactors may minimize these effects but they cannot be eliminated entirely or made negligible. The most important of these for mechanism evaluation include background offgasing of NOx and other reactive species, offgasing or heterogeneous reactions that cause “chamber radical sources” upon irradiation (e.g., see Carter and Lurmann, 1990, 1991), and ozone losses to the reactor walls. Most of these can be assessed by conducting various types of characterization experiments that either directly

W. P. L. Carter

32

measure the parameter of interest, or are highly sensitive to the chamber effect being assessed. The chamber effects that have been assessed and the types of experiments that have been utilized for assessing these effects in this chamber are summarized on Table 1 and discussed further below.

Table 1.

Summary of types of characterization experiments and types of chamber effects parameters derived from these experiments.

Run Type

No. Sensitive Runs Parameters

Comments

Ozone Dark Decay

4

O3 wall loss rate

The loss of O3 in the dark is attributed entirely to a unimolecular wall loss process.

CO - Air

8

NOx offgasing

Insensitive to radical source parameters but O3 formation is very sensitive to O3 offgasing rates. Formaldehyde data can also be used to derive formaldehyde offgasing rates.

CO - HCHO air

2

NOx offgasing.

Insensitive to radical source parameters but O3 formation is very sensitive to NOx offgasing rates. Also can be used to obtain formaldehyde photolysis rates

CO - NOx

6

Initial HONO, Radical source

O3 formation and NO oxidation rates are very sensitive to radical source but not sensitive to NOx offgasing parameters. Formaldehyde data can also be used to derive formaldehyde offgasing rates.

n-Butane - NOx

1

Initial HONO, Radical source

O3 formation and NO oxidation rates are very sensitive to radical source but not sensitive to NOx offgasing parameters.

Note that some of the chamber characterization parameters are derived by conducting model simulations of the appropriate characterization experiments to determine which parameter values best fit the data. All the characterization simulations discussed here were carried out using the SAPRC-99 chemical mechanism (Carter, 2000) with the photolysis and thermal rate constants calculated using the light characterization data, the measured temperatures of the experiments, and assuming no dilution for reasons discussed above. The rates of heterogeneous reactions not discussed below, such as N2O5 hydrolysis to HNO3 or NO2 hydrolysis to HONO, were derived or estimated based on laboratory studies or other considerations as discussed previously (e.g., Carter et al., 1995a). Although the assumed values of these parameters can affect model simulations under some conditions, they are not considered to be of primary importance in affecting simulations of the characterization or other experiments discussed here.

The UCR EPA Environmental Chamber

33

NOx offgasing NOx offgasing is the main factor limiting the utility of the chamber for conducting experiments under low NOx conditions. Although this can be derived by directly measuring increases in NOx species during experiments when NOx is not injected, the most sensitive measure is the formation of O3 in irradiations when VOCs but not NOx are initially present. The NOx offgasing rate is not determine directly, but derived by determining the NOx offgasing rates in model simulations of the experiments that fit the O3 yields obtained. The NOx offgasing can be represented in the model as inputs of any species that rapidly forms NOx in atmospheric irradiation systems, such as NO, NO2, or HONO (which rapidly photolyzes to form NO, along with OH radicals), but for reasons discussed below it is represented in our chamber effects model as offgasing of HONO, e.g., Walls + hQ o HONO

Rate = k1 x RN (1)

Where k1 is the light intensity as measured by the NO2 photolysis rate, and RN is the NOx (and radical) offgasing parameter, which is derived by model simulations of the appropriate characterization experiments to determine which value best fits the data. The NOx offgasing rates that fit the results of relevant experiments carried out in the first eight months of operation of this chamber are shown as the diamond symbols on Figure 3. It can be seen that the rates were around 1.5 ppt/min up to around run 85, then increased to 2-7 ppt/min after that, being somewhat higher in the “A” compared to the “B” reactor. The reason for this increase is unclear, but it may be related to the fact that maintenance was done to the reactors around the time of the change. The magnitudes of these apparent NOx offgasing rates are discussed further below.

NOx or Radical Input (ppt/min)

7 6 Radical Input (Side A)

5

Radical input (Side B)

4

NOx Input (Side A)

NOx Input (Side B)

3

Used in Model (Side A) 2

Used in Model (Side B)

1 0 50

75

100

125

150

175

EPA Run Number

Figure 3 .

Plots of apparent NOx or radical input rates against run number.

Chamber radical source It has been known for some time that environmental chamber experiments could not be modeled consistently unless some sources of radicals attributed to chamber effects is assumed (e.g., Carter et al., 1982; Carter and Lurmann, 1991; Carter, 2000). The most sensitive

34

W. P. L. Carter

experiments to this effect are NOx -air irradiations of compounds, such as CO or alkanes, which are not radical initiators and do not form radical initiating products to a sufficient extent to significantly affect their photooxidations. In some chambers at least part of the chamberdependent radical source can be attributed to formaldehyde offgasing (Simonaitis et al., 1997, Carter, 2004a), but as discussed below the magnitude of the formaldehyde offgasing in this chamber is not sufficient by itself for the model to simulate radical-source characterization experiments. For this chamber, assuming HONO offgasing at a similar magnitude as the apparent NOx offgasing rate derived as discussed above is usually sufficient to account for most of the chamber-dependent radical source, though results of some of the experiments are somewhat better simulated if a small amount (100 ppt or less) of HONO is also assumed to be initially present. The round symbols on Figure 3 shows plots of the HONO offgasing rates that are adjusted to fit the radical-source sensitive CO - NOx and n-butane - NOx experiments that were carried out in January-October of 2003. Note that since these experiments had initial NOx levels ranging from 10 - 200 ppb, they were not sensitive to NOx offgasing as such. However, from Figure 3 it can be seen that the magnitudes of the NOx offgasing and continuous radical input rates that fit the data for the respective characterization experiments were in the same range, and even changed at the same time when the characteristics of the chamber apparently changed. Whatever effect or contamination caused the apparent NOx offgasing to increase around the time of run 85 caused the same increase in the apparent radical source. Comparison of Radical Source and NOx Offgasing with Other Chambers Although HONO is not measured directly in our experiments, the fact that both the radical-sensitive and NOx-sensitive characterization experiments can be simulated assuming HONO offgasing at approximately the same rates is highly suggestive that this is the process responsible for both effects. Direct evidence for this comes from the data of Kleffmann et al. (2003), who used sensitive long path absorption photometer (LOPAP) instrument to detect ppt levels of HONO emitted from the walls during irradiations in the large outdoor SAPHIR (Brauers et al, 2003) at rates comparable to those observed in the earlier experiments in this chamber. The SAPHIR chamber is similar in design to the chamber discussed here, except it is larger in volume and is located outdoors. In particular, like our chamber it has Teflon walls and uses a dual-reactor configuration to minimize contamination by outside air. Therefore, it would be expected to have similar chamber NOx and radical sources, and this appears to be the case. Figure 4 shows plots of the NOx offgasing or radical source parameter (e.g. RN in Equation 1) obtained in modeling appropriate characterization runs in various chambers, where they are compared with direct measurements made in the SAPHIR chamber (Kleffmann et al. (2003). In addition to this UCR EPA chamber, the radical source parameters shown are those derived by Carter (2000) for previous indoor and outdoor chambers at UCR (Carter et al., 1995a), those derived by Carter and Lurmann (1990, 1991) for the University of North Carolina (UNC) outdoor chamber (e,g., Jeffries et al., 1982, 1990), and those derived by Carter (2004a) for the Tennessee Valley Authority (TVA) indoor chamber (Simonaitis and Bailey, 1995; Bailey et al., 1996). (Note that the data shown for the UCR EPA chamber includes experiments carried out subsequently to those shown on Figure 3, including a few runs at reduced temperature.) The figure shows that the radical source and NOx offgasing rates derived for this chamber are comparable in magnitude to the HONO offgasing directly

The UCR EPA Environmental Chamber

35

measured in the SAPHIR chamber and also comparable to the NOx offgasing derived for TVA chamber but are significantly lower than those derived from modeling characterization data from the earlier UCR and UNC chambers. It is interesting to note that parameters derived for the various chambers all indicate that the radical source and HONO or NOx offgasing rates all increase with temperature.

1000 NOx offgasing or Radical Source / NO2 photolysis rate (ppt)

Previous UCR Indoor Previous UCR Outdoor

Previous UCR Default Model 100

UNC Outdoor Model

TVA Indoor (NOx offgasing) UCR EPA (first series)

10

UCR EPA (later series) SAPHIR HONO offgasing (dry) 1 290

300

310

320

Average Temperature (K)

Figure 4 .

Plots of ratios of NOx offgasing or radical source to the NO2 photolysis rates derived from modelling characterization runs for various chambers.

Therefore, the radical source and NOx offgasing rates indicated by the characterization data for the first series of experiments for this chamber is probably as low as one can obtain for reactors constructed of FEP Teflon film, which is generally believed to be the most inert material that is practical for use as chamber walls. Although the radical source and NOx offgasing rates for the second series of experiments is higher (see also Figure 3), they are still about an order of magnitude lower than observed for the UCR and UNC chambers previously used for mechanism evaluation. Formaldehyde offgasing Low but measurable amounts of formaldehyde were formed in irradiations in this chamber, even in pure air, CO - NOx, or other experiments where no formaldehyde or formaldehyde precursors were injected. Some formaldehyde formation is expected from the reactions of methane in the matrix air, since the methane removal catalyst was not operational during the period of these experiments, but formaldehyde formation in this reaction in the CO experiments is predicted to be negligible. The data in essentially all such experiments could be modelled assuming a continuous light-dependent formaldehyde offgasing rate corresponding to 7.5 ppb/day at the light intensity of these experiments. This is a relatively low offgasing rate that could not be detected with formaldehyde analyzers used in most previous UCR and other chamber experiments, and is insufficient to account for the apparent chamber radical

36

W. P. L. Carter

source observed in most chamber experiments. This apparent formaldehyde offgasing has a non-negligible effect on very low VOC and radical source characterization experiments, so it must be included in the characterization model. However, it has a relatively minor impact on modelling most experiments used for VOC mechanism evaluation or reactivity assessment. The source of the apparent formaldehyde offgasing in the Teflon reactors is unknown, but it is unlikely to be due to build-up of contaminants from previous exposures or contamination from the enclosure. The apparent formaldehyde offgasing rate is quite consistent in most cases and there are no measurable differences between the two reactors or changes in formaldehyde offgasing with time. The data are best modelled by assuming only direct formaldehyde offgasing, as opposed to some formaldehyde precursor being formed from light-induced reactions of some undetected contaminant, as is apparently important in the TVA chamber (Simonaitis et al., 1997; Carter, 2004a). Representative Results of Initial Evaluation Experiments Table 2 gives a summary of the initial experiments carried out in this chamber for chamber and mechanism evaluation. The characterization and single organic - NOx experiments and some of the ambient reactive organic gas (ROG) surrogate - NOx experiments were funded by the EPA cooperative agreement used to fund the construction and initial characterization of this chamber. The lowest concentration ambient ROG surrogate NOx experiments were carried out by a separate California Air Resources Board project to obtain additional data for low NOx mechanism evaluation, and a number of surrogate - NOx experiments at varying initial ROG and NOx levels at a were carried out for a separate program to obtain data to evaluate observationally based methods. Not shown on Table 2 are a number of “incremental reactivity” experiments, where various compounds or VOC mixtures were added to standard ambient ROG - NOx surrogate irradiations to provide data to test mechanisms for their relative ozone impacts. The results of these experiments are still being analyzed. The various characterization experiments were used to derive the chamber characterization parameters and evaluate the chamber characterization model as discussed above. The single organic - NOx experiments were carried out to demonstrate the utility of the chamber to test the mechanisms for these compounds, for which data are available in other chambers, and to obtain well-characterized mechanism evaluation data at lower NOx levels than previously available. The formaldehyde + CO - NOx experiments were carried out because they provided the most chemically simple system that model calculations indicated was insensitive to chamber effects, to provide a test for both the basic mechanism and the light characterization assignments. The aromatic - CO - NOx experiments were carried out because aromatic - NOx experiments were predicted to be very sensitive to the addition of CO, because it enhances the effects of radicals formed in the aromatic system on ozone formation. The ambient surrogate - NOx experiments were carried out to test the ability of the mechanism to simulate ozone formation under simulated ambient conditions at various ROG and NOx levels.

The UCR EPA Environmental Chamber

Table 2.

37

Summary of initial chamber and mechanism evaluation experiments.

Runs

NOx Range (ppb)

VOC Range (ppm)

Avg. Model Bias #

Avg. Model Error #

Characterization

32

0 - 200

Varied

-3%

28%

Formaldehyde - NOx

2

8 - 25

0.35 – 0.50

-23%

23%

Formaldehyde - CO - NOx

2

15 - 20

HCHO: 0.4 – 0.5 CO: 15 - 80

-10%

10%

Ethene - NOx

2

10 - 25

~0.6

-15%

15%

Propene - NOx

2

5 - 25

0.4 – 0.5

16%

16%

Toluene or m-Xylene - NOx

4

5 - 25

10%

10%

Aromatic - NOx + CO

6

5 - 30

Toluene: 0.6 – 0.16 m-Xylene: 0.18 CO: 25 – 50

-17%

18%

61*

2 - 315

0.2 – 4.2 ppmC

-10%

13%

Run Type

Ambient Surrogate - NOx

* Includes experiments carried out for subsequent programs. # Error and bias for model predictions of '([O3]-[NO]) using the SAPRC-99 mechanism. Bias is (calculated - experimental) / calculated. Error is the absolute value of the bias.

The ROG surrogate used in the ambient surrogate - NOx experiments consisted of a simplified mixture of designed to represent the major classes of hydrocarbons and aldehydes measured in ambient urban atmospheres, with one compound used to represent each model species used in condensed lumped-molecule mechanism. The eight representative compounds used were n-butane, n-octane, ethene, propene, trans-2-butene, toluene, m-xylene, and formaldehyde. (See Carter et al., 1995b, for a discussion of the derivation of this surrogate). The ability of the SAPRC-99 mechanism to simulate the total amount of NO oxidized and O3 formed in the experiments, measured by ([O3]final-[NO]final) - ([O3]initial-[NO]initial) or '([O3]-[NO]), is summarized for the various types of experiments on Table 2 and shown for the individual runs on Figure 5. This gives an indication of the biases and run-to-run variability of the mechanism in simulating ozone formation. Note that in experiments with excess NO the processes responsible for O3 formation are manifested by consumption of NO, so simulations of '([O3]-[NO]) provides a test of model simulations of these processes even for experiments where O3 is not formed. Note that the characterization runs were modelled using the default characterization parameters used when modelling the mechanism evaluation runs, not with the values that were adjusted to fit the individual experiments. Therefore, the relatively large variability and average model error for the model simulations of '([O3]-[NO]) in those experiments provides a measure of the variability of the chamber effects parameters (e.g., HONO offgasing) to which these experiments are sensitive. The relatively low average bias is expected because the chamber effects parameter values were derived based on these data.

W. P. L. Carter

38 Single Run

Average

All Characterization HCHO - NOx HCHO - CO - NOx Ethene - NOx Propene - NOx Toluene - NOx Toluene - CO - NOx m-Xylene - NOx m-Xylene - CO - NOx Surrogate - NOx -75%

-50%

-25%

0%

25%

50%

75%

(Calculated - Experimental) / Experimental

Figure 5 .

Fits of experimental '([O3]-[NO]) measurements to SAPRC-99 model calculations for the initial chamber and mechanism evaluation experiments.

For the single VOC - NOx or VOC - CO - NOx experiments, the model is able to simulate the '([O3]-[NO]) to within r25% or better in most cases, which is better than the r~30% seen in previous mechanism evaluations with the older chamber data (Carter and Lurmann, 1990, 1991; Gery et al., 1989, Carter, 2000). However, there are indications of nonnegligible biases in model simulations of certain classes of experiments. The cleaner conditions and the relatively lower magnitude of the chamber effects may make the run-to-run scatter in the model performance may be less than in the simulations of the previous data, and this tends to make relatively small biases in the model performance more evident. There are, for example, definite biases in the model to under predict O3 formation and NO oxidation in the surrogate - NOx experiments carried out at lower ROG/NOx ratios. These cases are discussed further elsewhere (Carter, 2004b). As indicated on Table 2, the initial evaluation experiments included runs with NOx levels as low as 2-5 ppb, which is considerably lower than in experiments used previously for mechanism evaluation. Most of the experiments used in the previous SAPRC-99 mechanism evaluation had NOx levels greater than 50 ppb, and even the “low NOx” TVA and CSIRO experiments had NOx levels of ~20 ppb or greater, except for a few characterization runs (Carter, 2004a, and references therein). However, there is no indication in any difference in model performance in simulating the results of these very low NOx experiments, compared to those with the higher NOx levels more representative of those used in the previous evaluation. For example, Figure 6 shows concentration-time plots for selected measured species in ambient surrogate - NOx experiment carried out at the lowest NOx levels in the initial evaluation runs. To indicate the sensitivity of the experiments to NOx offgasing effects, the effects of varying the HONO offgasing parameter from zero to the maximum level consistent with the characterization experiments is also shown. It can be seen that the model using the default HONO offgasing parameter value gives very good fits to the data. Although the O3 simulations are somewhat affected when the HONO offgasing rate is varied within this somewhat extreme range, the sensitivity is not so great that the uncertainty in this parameter

The UCR EPA Environmental Chamber

39

significantly affects conclusions one can draw about the ability of the model to simulate this low NOx experiment. However, the sensitivity would increase as the NOx levels are reduced, and ~2 ppb NOx probably represents a reasonable lower limit for NOx levels useful for mechanism evaluation. O3

0.04

NO

0.002

NO2

0.0015

0.03

0.0010

0.001

0.02

0.0005

0.01 0.00

0.000

0

120

240

360

m-Xylene

0.0000

0

120

240

360

120

240

360

240

360

PAN

NOy - HNO3

0.004

0

0.0006

0.006

0.003 0.004

0.0004

0.002

0.002

0.0002

0.001 0.000

0.000 0

120

240

360

0.0000

0

120

240

360

0

120

Concentration (ppm) vs Time (minutes)

Experimental No HONO Offgasing

Figure 6.

Standard Model Maximum HONO Offgasing

Concentration-time plots of selected compounds in the lowest NOx ambient ROG - NOx surrogate experiment in the initial evaluation experiments (NOx | 1 ppb, ROG | 300 ppbC.

Overall, the results of the initial characterization and evaluation indicate that this chamber can provide high quality mechanism evaluation data for experiments with NOx levels as low as ~2 ppb, considerably lower than employed in previous experiments. Chamber effects are not absent, but they are as low as or lower than in observed in any previous chambers used for mechanism evaluation, in some cases by an order of magnitude or more. Although a larger number of experiments would be required to fully assess this, the results also suggest a higher degree of precision in mechanism evaluation than observed previously, making smaller biases in mechanism performance more evident. The initial dataset from this chamber indicate no significant problems with mechanism performance that are characteristic of low NOx conditions as such, but do reveal problems with the mechanisms for aromatics and the ambient ROG surrogate. This is discussed further in our companion presentation (Carter, 2004b). Future Research Directions The future research directions for the UCR EPA chamber will of course depend on funding availability and the interests of the funding agencies, but we would expect the research would be designed to take full advantage of its capabilities. In any case, we expect to continue the ozone reactivity and gas-phase mechanism studies that are currently underway, with the mechanism evaluation being focused on aromatics, whose mechanism are the most

40

W. P. L. Carter

uncertain and where the new data indicate mechanism problems (Carter, 2004a,b), and ozone reactivities of VOC source categories of interest. We will also be utilizing the chambers for initial experiments to investigate temperature effects on O3 formation and other measures of gas-phase reactivity, taking advantage of the temperature control capabilities of this chamber, which has not yet been fully exploited. We hope in the future to obtain instrumentation needed for NO3, N2O5, HOx, and other trace species in order to improve capabilities and utility of this facility for mechanism evaluation. As discussed in more detail by Cocker (2004), we will also utilize the chamber for well-characterized experiments needed to develop and evaluate models for secondary organic aerosol (SOA) formation. The temperature control capabilities of this chamber permit systematic studies of temperature and humidity effects that are not possible with existing outdoor chambers or with indoor chambers using blacklights (whose intensities are temperature-dependent). The established capability of this chamber for gas phase mechanism evaluation is an important characteristic in this regard, since the model cannot properly simulate SOA formation if it does not also properly simulate the formation of the SOA precursors from the gas-phase reactions. To be fully utilized, this chamber should also serve as a resource for collaborative studies where environmental chamber measurements under highly controlled conditions would be useful. We encourage collaboration with other researchers who have instrumentation that we lack that would be useful for various studies, such as our collaboration with William Brune of Penn State University, whose HO and HO2 radical measurement instrumentation was used with the surrogate - NOx experiments carried out for the observational based methods study. This facility would also be useful as a test bed for evaluating analytical instrumentation for use in ambient monitoring, since it can produce realistic but highly characterized simulated pollution conditions for instrument evaluation and inter-comparisons. References Bailey, E. M., C. H. Copeland and R. Simonaitis; Smog chamber studies at low VOC and NOx concentrations, Report on Interagency Agreement DW64936024 to EPA/NREL, Research Triangle Park, NC. (1996). Brauers, T., B. Bohn, F.-J. Johnen, F. Rohrer, S. Rodriguez Bares, R. Tillmann, and A. Wahner; The atmosphere simulation chamber SAPHIR: a tool for the investigation of photochemistry, Presented at the EGS AGU - EUG Joint Assembly, Nice, France, April 11 (2003). See also http://www.fz-juelich.de/icg/icgii/saphir/home. Carter, W. P. L.; Documentation of the SAPRC-99 chemical mechanism for VOC reactivity assessment, Report to the California Air Resources Board, Contracts 92-329 and 95-308, May 8 (2000). Available at http://www.cert.ucr.edu/~carter/absts.htm#saprc99. Carter, W. P. L.; Development of a next-generation environmental chamber facility for chemical mechanism and VOC reactivity research. Draft research plan and first progress report to the United States Environmental Protection Agency Cooperative Agreement CR 827331-01-0. January 3 (2002). Available at http://www.cert.ucr.edu/~carter/epacham. Carter, W. P. L.; Evaluation of a gas-phase atmospheric reaction mechanism for low NOx conditions, Final Report to California Air Resources Board Contract No. 01-305, May 5 (2004a). Available at http://www.cert.ucr.edu/~carter/absts.htm#lnoxrptf Carter, W. P. L.; Environmental chamber studies of ozone formation potentials of VOCs. Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004b).

The UCR EPA Environmental Chamber

41

Carter, W. P. L., R. Atkinson, A. M. Winer, and J. N. Pitts, Jr.; Experimental Investigation of ChamberDependent Radical Sources, Int. J. Chem. Kinet., 14 (1982) 1071. Carter, W. P. L., and F. W. Lurmann; Evaluation of the RADM gas-phase chemical mechanism. Final Report to the U.S. Environmental Protection Agency, EPA-600/3-90-001 (1990). Carter, W. P .L., and F. W. Lurmann; Evaluation of a detailed gas-phase atmospheric reaction mechanism using environmental chamber data. Atmos. Environ. 25A (1991) 2771-2806. Carter, W. P. L., D. Luo, I. L. Malkina, and D. Fitz; The University of California, Riverside environmental chamber data base for evaluating oxidant mechanism. Indoor chamber experiments through 1993, Report submitted to the U. S. Environmental Protection Agency, EPA/AREAL, Research Triangle Park, NC., March 20 (1995a). Available at http://www.cert.ucr.edu/~carter/ absts.htm#databas. Carter, W. P. L., D. Luo, I. L. Malkina, and J. A. Pierce; Environmental chamber studies of atmospheric reactivities of volatile organic compounds. Effects of varying ROG surrogate and NOx, Final report to Coordinating Research Council, Inc., Project ME-9, California Air Resources Board, Contract A0320692, and South Coast Air Quality Management District, Contract C91323. March 24 (1995b). Available at http://www.cert.ucr.edu/~carter/absts.htm#rct2rept. Cocker, D. R. State-of-the art chamber for aerosol studies. Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004). Dodge, M. C.; Chemical oxidant mechanisms for air quality modeling, critical review paper for 1998 Ozone Assessment, Atmos. Environ, 34 (2000) 2103-2130. Gaffney, J. S, R. M. Bornick., Y-H Chen, and N. A. Marley; Capillary gas chromatographic analysis of nitrogen dioxide and PANs with luminol chemiluminescent detection, Atmos. Environ. 32 (1998) 1445-1454. Gery, M. W., G. Z. Whitten, J. P. Killus, and M. C. Dodge; A Photochemical Mechanism for Urban and Regional Scale Computer Modeling, , J. Geophys. Res, 94, (1989) 12,925. Hastie, D. R, G. I. Mackay, T. Iguchi, B. A. Ridley, and H. I. Schiff, H. I; Tunable diode laser systems for measuring trace gases in tropospheric air, Environ. Sci. Technol. 17, (1983) 352A-364A. Jeffries, H. E., R. M. Kamens, K. G. Sexton, and A. A. Gerhardt; outdoor smog chamber experiments to test photochemical models, EPA-600/3-82-016a, April (1982). Jeffries, H. E., K. G. Sexton, J. R. Arnold, Y. Bai, J. L. Li, and R. Crouse; A Chamber and Modeling Study to Assess the Photochemistry of Formaldehyde, Report on EPA Cooperative Agreement CR-813964, Atmospheric Research and Exposure Assessment Laboratory, EPA, Research Triangle Park, NC. (1990). Jeffries, H.E.; M. W. Gery, and W. P. L. Carter; Protocol for evaluating oxidant mechanisms for urban and regional models. Report for U.S. Environmental Protection Agency Cooperative Agreement No. 815779, Atmospheric Research and Exposure Assessment Laboratory, Research Triangle Park, NC. (1992). Kleffmann, J., B. Bohn, T. Brauers, F. Rohrer, A. Wahner, and P. Wiesen: Photolytic HONO formation in the atmospheric simulation chamber SAPHIR. Presented at the EGS - AGU - EUG Joint Assembly, Nice, France, April 11 (2003). Schiff, H.. I., G. I. Mackay, and J. Bechara; The use of tunable diode laser absorption spectroscopy for atmospheric measurements, Res. Chem. Intermed. 20, 1994 (1994) 525-556. Simonaitis, R. and E. M. Bailey; Smog chamber studies at low VOC and NOx concentrations: Phase I, Report on Interagency Agreement DW64936024 to EPA/NREL, Research Triangle Park, NC. (1995). Simonaitis, R., J. Meagher, and E. M. Bailey; Evaluation of the condensed Carbon Bond Mechanism against smog chamber data at low VOC and NOx Concentrations, Atm. Environ. 31 (1997) 27-43. Zafonte, L., P. L. Rieger, and J. R. Holmes; Nitrogen dioxide photolysis in the Los Angeles atmosphere, Environ. Sci. Technol. 11 (1977) 483-487.

Investigations of Secondary Organic Aerosol in the UCR EPA Environmental Chamber David R. Cocker III and Chen Song Department of Chemical and Environmental Engineering, UC Riverside, Riverside, California, 92521, USA Key Words: Environmental chambers, Secondary organic aerosol, Oxides of nitrogen, m-Xylene

Abstract A large indoor environmental reactor was completed in 2002 at UC Riverside to investigate photochemical processes leading to ozone and secondary organic aerosol (SOA) formation. The chamber is briefly described by Carter (2004) in a separate contribution to this workshop and in a more detailed manuscript submitted to Atmospheric Environment (Carter et al., 2005). The contribution to the workshop covered some details of the UCR chamber wall characterization (Carter, 2004), selected approaches to studies of secondary organic aerosol formation (Cocker et al., 2001 and Song et al., 2005), the NOx dependence of m-xylene SOA formation (Song et al, 2005), some acid catalyzed reaction work from Caltech (Gao et al., 2004), and select details of the Caltech environmental chamber (Cocker et al., 2001). This contribution will briefly describe the SOA components of the chamber facility and advantages of the new reactor with respect to SOA investigations, and will summarize the findings on the importance of hydrocarbon:NOx ratio on SOA formation in the m-xylene photooxidation system. Introduction The environmental chamber can be considered to be a well-mixed batch reactor where one can study atmospheric processes in the absence of microphysical transport processes. Volumes for such reactors range from tens of liters to hundreds of cubic meters with a variety of surfaces selected to minimize the impact of the walls on the chemistry occurring within the reactor (Becker, 2004 and references therein). Secondary organic aerosol remains a significant contributor to regional air pollution with estimates of up to 80% of fine particulate being secondary in origin in heavily impacted regions (Turpin et al., 1991 and Na et al., 2004) and may play an important role in visibility degradation, global climate change, and adversely affect human health (Seinfeld, 2004 and references therein). Traditionally, the extent of secondary organic aerosol formation from a reacted hydrocarbon precursor has been discussed in terms of aerosol yield (Y), defined as the mass concentration of aerosol formed divided by the mass concentration of reacted hydrocarbons. The aerosol yield is then described using a two-parameter semi-empirical approach (Pankow, 1994a,b and Odum et al., 1996 ) Y

¦Y

i

i

'M org ¦ i

D i Ȁ om,i

(1)

1  Ȁ om,i 'M org 43

I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 43–47. © 2006 Springer. Printed in the Netherlands.

D. R. Cocker III and C. Song

44

where Yi is the yield of individual aerosol forming products, Di is the mass-based stoichiometric fraction of species i formed from the parent hydrocarbon, Kom,i is the gasparticle partitioning coefficient (m3Pg-1), which is inversely proportional to the compound’s vapor pressure, and 'Morg (Pgm-3) is the total mass concentration of organic material and associated water present in the aerosol phase. The fraction of secondary organic material condensing into the aerosol phase is seen to depend on the amount of organic aerosol mass present. Furthermore, Kom,i can be estimated (Pankow, 1994a,b) from K om ,i

760 RTf om ,i 10 MWom] i p Lo ,i 6

(2)

where R is the ideal gas constant; T is temperature; fom is the weight fraction of the total suspended particulate that comprises the absorbing phase; MWom is the number-averaged molecular weight of the absorbing organic matter phase, ]i is the activity coefficient of species i and poL,i is the sub-cooled liquid vapor pressure. The two-product semi-empirical model then assumes that two surrogate species can be used to estimate the SOA yield: one surrogate product representing low vapor pressure compounds and one surrogate product representing high vapor pressure compounds (i=1,2 in Equation 1). Current application of the two-product model does not provide an obvious means to incorporate changes in gas-phase chemistry due to the existence of nitrogenous species, light hydrocarbons, or water. Little work has been performed to date on the effect of NOx on SOA yield. Izumi et al. (1990) performed two experiments with the same initial m-xylene concentrations but different initial NOx concentrations and found similar SOA yields. Odum et al. (1996) investigated SOA formation from m-xylene photooxidation and concluded that the SOA yield for aromatic hydrocarbons is weakly dependent on HC:NOx. Hurley et al. (2001) investigated SOA formation from toluene and concluded that SOA yield for toluene with and without ozone (high NOx) was 0.108±0.004 and 0.075±0.004, respectively. Most recently, Wirtz (2004) at this conference and Martins-Reviejo et al. (2005) reported increased aerosol formation per reacted hydrocarbon at lower NOx concentrations for benzene and toluene. This paper (published in detail by Song et al. (2005) presents significantly increased aerosol formation per reacted m-xylene as hydrocarbon to NOx ratio increases. Also presented are some details on current aerosol measurement capabilities and unique advantages for SOA work in this facility. Brief chamber description The UC Riverside facility consists of dual 90 m3 chambers attached to a movable frame that is slowly lowered to maintain a slightly positive differential pressure. The chambers are located within a large, temperature controlled refrigeration container that is continually flushed with purified dry air. Two light sources are available: a 200 kW argon arc lamp and two banks of 40 each 72 W 1.22 m fluorescent blacklamps. The temperature and humidity within the chamber can be accurately controlled (See Carter et al., 2005 for more detail). Advantages of the indoor reactor include temperature (recall from equation 2 the

Investigations of Secondary Organic Aerosol in the UCR EPA Environmental Chamber

45

importance of temperature on Kom,i) and humidity control. Additionally, the indoor reactors benefit from reproducible lighting conditions throughout an experiment and experiment to experiment and provide for improved comparison between different experiments. However, the key disadvantage is the lack of a true “solar” radiation spectrum. Spectral differences are largely overcome in this reactor due to a 200 kW argon arc lamp irradiation source that performs a reasonable simulation of the UV-visible outdoor solar spectra (see Carter, 2004). Regardless of the location of the chamber (indoor vs. outdoor), the absorption properties of the film will lead to some modification of incoming radiation. Instrumentation: Basic aerosol instrumentation is currently limited to a suite of Scanning Electrical Mobility Spectrometers (SEMS) (Wang and Flagan, 1989) that monitor particles in the 28-730 nm range. The instrument design and operation is very similar to those described by Cocker et al. (2001). Sheath air for particle classification is drawn and filtered from the atmospheric reactor to ensure that the sheath gas surrounding the particles would be similar in chemical composition thus reducing bias due to evaporation or condensation of semi-volatile material. The classification portion of the SEMS are housed within the temperature controlled enclosure to ensure that no bias occurs in sampling due to changes in sampling line temperatures. Additionally, a fast scanning mobility particle spectrometer (f-SMPS) (size range 3-100 nm, 1 second time resolution, see also Shah et al., 2005) and a tandem differential mobility analyzer (TDMA, see also Cocker et al., 2001) are available to track ultrafine particle production and aerosol hygroscopic properties, respectively. Offline instrumentation for detailed chemical characterization of gaseous, semi-volatile, and particulate products are also available. Particle loss rate: The particle loss rate for the chamber can be described as a first-order wall loss process (Cocker et al., 2001). Particle loss rates determined for the first set of aerosol experiments within the reactor is about 7 day-1. While maximum particle volume in experiments performed have ranged from less than 0.1 to almost 80 µg/m3, no correlation between maximum particle volumes and measured decay rates has been seen (Carter et al., 2005). Impact of NOx on aerosol formation A series of experiments have been performed in the UCR chamber to ascertain impacts of NOx on aerosol formation in the m-xylene/NOx experiments. A full research paper on this subject has already been published in ES&T (Song et al.) A brief summary of the work (presented in Zakopane) follows. A series of 24 m-xylene photoxidation experiments (blacklamp irradiation) were performed at 300K. Reacted hydrocarbon mass concentrations ranged from 149 to 1660 Pg m-3, NOx concentrations ranged from 9 to 587 ppb, and HC:NOx ratios (ppbC/ppb NOx) ranged from 1.3 to nearly 50. All experiments were performed until wall loss corrected aerosol volume plateaued reached a constant stable value. While the data for all experiments could be plotted as yield versus organic mass produced and generally explained by a two-product fit (0.040, 0.164, 0.250 and 0.012 for Į1,

D. R. Cocker III and C. Song

46

Į2, K1 and K2), closer inspection of the aerosol mass concentration produced for experiments with similarly reacted hydrocarbons showed significantly different total aerosol concentration. The discrepancy between quality of fit for Y vs. Mo and the realization that aerosol mass concentration produced varied significantly for similar hydrocarbon reacted was simply attributed to the fact that the y-axis was not truly independent of the x-axis. (Y is inversely proportional to Mo; Mo is the x-axis variable). Therefore, to see the effect of NOx on SOA formation more clearly, the data was replotted as Mo versus HC reacted and the two-product model solved as Mo

1 ª 2 2 K1K2'HCD1  D2  K1  K2  K12 K22'HC2 D1  D2  2K1K2'HCK1  K2 D1  D2  K1  K2 º »¼ 2K1K2 «¬

(3)

This allowed for the determination of two sets of distinct aerosol curves, one for low HC:NOx ratios (<5.5) and one for high HC:NOx ratios (>8.0). Values of Į1, Į2, K1 and K2 are 0.024, 0.152, 0.229, 0.004 for low HCo:NOx ratio experiments (R2>0.99) and 0.049, 0.178, 0.301, 0.008 for high HCo:NOx ratios (R2>0.99). (Song et al., 2005). A plot of the modeled aerosol mass concentrations versus hydrocarbon reacted can be found in Figure 1.

400 HCo:NOx<5.5 HCo:NOx>8.0

350

Mo(Pg/m3)

300 250 200 150 100 50 0 0

500

1000

1500

2000

3

HC(Pg/m ) Figure 1.

SOA mass concentrations from m-xylene photooxidation as a function of mxylene consumption. Empirical fit using equation 3 for low NOx (HCo:NOx>8.0) and high NOx (HCo:NOx<5.5) conditions.

Of critical interest is the large deviation in aerosol produced between the low NOx (conditions more typical of the ambient) and high NOx (conditions often employed in atmospheric reactors) conditions, particularly for aerosol mass concentrations and reacted hydrocarbon concentrations typical of the polluted urban atmosphere (lower left portion of the curve).

Investigations of Secondary Organic Aerosol in the UCR EPA Environmental Chamber

47

Finally, it must be noted that the discrepancy between the aerosol productions could not be attributed solely to total nitrate radical or ozone concentrations. Furthermore, the onset of aerosol formation (aerosol threshold) in the systems was found to be strongly dependent on the HC:NOx ratios present in the chamber. References Becker, K.H. Historical Overview on the Development of Chambers for the Study of Atmospheric Processes; Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004). Carter, W.P.L. The CE-CERT Chamber in Riverside, California; Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004). Carter, W.P.L., D.R. Cocker III, D.R. Fitz, I.L. Malkina, K. Bumiller, C.G. Sauer, J.T. Pisano, C. Bufalino, C. Song; A New Environmental Chamber for Evaluation of Gas-Phase Chemical Mechanisms and Secondary Aerosol Formation. Submitted to Atmospheric Environment, 2005. Cocker D.R. ɒ, R.C. Flagan, J.H. Seinfeld; State-of-the-art chamber facility for studying atmospheric aerosol chemistry. Envir. Sci. Technol. 35 (2001) 2594-2601. Gao S., N.L. Ng, M. Keywood, V. Varutbangkul, R. Bahreini, A. Nenes, J.W. He, K.Y. Yoo, J.L. Beauchamp. R.P. Hodyss, R.C. Flagan, J.H. Seinfeld; Particle phase acidity and oligomer formation in secondary organic aerosol. Envir. Sci. Technol. 38 (2004) 6582-6589. Hurley M. D., O. Sokolov, T.J. Wallington, H. Takekawa, M. Karasawa, B. Klotz, I. Barnes, and K. Becker; Organic aerosol formation during the atmospheric degradation of toluene. Environ. Sci. Technol. 35 (2001) 1358-1366. Izumi K. and T. Fukuyama; Photochemical aerosol formation from aromatic-hydrocarbons in the presence of NOx. Atmos. Environ. 24A:6 (1990), 1433-1441. Martin-Reviejo, M. and K. Wirtz. Is Benzene a Precursor for Secondary Organic Aerosol? Environ. Sci. Technol.39 (2005), 1045-1054. Na K., A.A. Sawant, C. Song, and D.R. Cocker; Primary and secondary carbonaceous species in the atmosphere of Western Riverside County, California. Atmos. Environ. 38 (2004) 1345-1355. Odum J. R., Hoffmann T., Bowman F., Collins D., Flagan R. C. and Seinfeld J. H. Gas/particle partitioning and secondary organic aerosol yields. Envir. Sci. Technol. 1996, 30, 2580-85. Pankow J. F. An absorption-model of gas-particle partitioning of organic-compounds in the atmosphere. Atmos. Environ. 1994a, Vol. 28, No. 2, 185-188. Pankow J. F. An absorption model of the gas-aerosol partitioning involved in the formation of secondary organic aerosol. Atmos. Environ.1994b, Vol. 28, No. 2, 189-193. Seinfeld, J.H, and Pankow, J.F. Organic Atmospheric Particulate Material. Annu. Rev. Phys. Chem.¸2003, Vol. 54:121-140. Shah, S.D., and Cocker III,D.R., A Fast Scanning Mobility Particle Spectrometer for Monitoring Transient Particle Size Distributions, submitted to Aero. Sci. & Technol., 2005. Song C., Na K., Cocker D. R. Impact of Hydrocarbon to NOx ratio on Secondary Organic Aerosol Formation. Envir. Sci. Technol. 2005, In Press. Turpin B. J. and Huntzicker J. J. Secondary formation of organic aerosol in the Los-angles basin – a descriptive analysis of organic and elemental carbon concentrations. Atmos. Environ. 1991, Vol. 25A, No. 2, 207215. Wang S. C., Flagan R. C. Scanning Electrical Mobility Spectrometer. Aero. Sci. Technol. 1990, 13, 230-240. Wirtz, K. Recent Investigations on Aerosol Formation in EUPHORE Chamber. Presented at the NATO ESTARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004).

Field Measurement and Modelling Motivated Design of a Programme of Aerosol Chamber Experiments Gordon McFiggans Atmospheric Sciences Group, School of Earth, Atmospheric & Environmental Sciences (SEAES), University of Manchester, Sackville Street Building, Manchester, M60 1QD, UK Key Words: Atmospheric aerosol, Chamber studies, Coupled modelling

Abstract This paper describes the development of a chamber facility designed to investigate a range of atmospheric aerosol processes driven by their potential to affect radiative forcing. Development of a suite of models investigating coupled chemical and microphysical processes has informed the definition of a preliminary experimental programme. The chamber construction is underway and the experiments are phased to follow the chamber development and completion. Introduction Atmospheric aerosols are the focus of vigorous and extensive current investigation due to their roles in climate change through direct and indirect radiative forcing and to their impacts on human health. This paper focuses on the development of a chamber facility to investigate aerosols and aerosol processes which may impact on both, but will be characterised in terms of their ability to participate in climate change. The climatic implications depend critically on the number of aerosols of optically active size, their composition and resulting affinity for water in the sub-saturated environment and their ability to activate into cloud droplets and ice crystals moving into supersaturation. Using climatic implications as primary development drivers it was decided to design an aerosol chamber and work programme of targeted experiments to investigate example aerosol systems and to demonstrate the suitability of such a chamber to probe the changes in radiative forcing potential with aerosol processing. In contrast to atmospheric gas phase processes, the understanding of the cycling and implications of atmospheric aerosols is relatively young and underdeveloped. This is, in part, due to the complexity of the coupling between the microphysics and chemistry and also due to the lack of availability of suitable instrumentation to probe aerosol chemical composition and physical behaviour in the field with sufficient time resolution (and even in the laboratory with sufficient selectivity). Given the current state-of-the-science, it is inevitable that field observations of aerosol physical and chemical properties regularly throw up surprises and unanswered questions. It is unclear whether the most important questions are being addressed in current detailed laboratory studies where many of the approaches have been borrowed from the gas phase laboratory work. Recent adaptation and application of techniques from surface chemistry show promise in terms of detailed mechanistic interpretation of multiphase processes, but the interpretation of findings from such studies and their incorporation into even simple atmospheric models is difficult (due to problems extrapolating from aerosol mimics, for example). In addition, the degree of complexity in gaseous composition has been widely reported (e.g. Lewis et al., 2000) and, though less well characterised, it is likely that the composition of aerosols will exhibit comparable complexity. Indeed, due to the range of 49 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 49–65. © 2006 Springer. Printed in the Netherlands.

50

G. McFiggans

sizes and possible mixing states, a detailed understanding of the intricate chemical and microphysical couplings in many aerosol systems is unlikely to be tractable, and may not even be desirable. Whilst modern kinetic flow studies are needed to understand underlying mechanistic concepts and levitation techniques can clarify single particle thermodynamic behaviour, it is difficult to directly and quantitatively evaluate the importance of such complex systems by testing models using such data directly against field measurements. Chamber facilities have played a central role in recent developments in gas phase atmospheric chemistry, allowing the development of understanding of complex mechanisms and they are increasingly used to investigate the formation and evolution of aerosols. The main advantage of chamber studies is the ability to separate chemical and microphysical atmospheric processes from the effects of large scale flow and mixing and it is for studies of continuous progressive evolution of atmospheric components that the use of a chamber as a well-mixed and closed batch reactor is suited. An additional important advantage of chamber studies is the ability to use relatively representative reactant concentrations and to study processes taking place over realistic atmospheric timescales. The reduced surface area to volume ratios with increasing reactor size enables this. A typical atmospherically relevant aerosol evolution timescale might be of the order of several hours to days. Therefore, chamber studies investigating the microphysical and chemical evolution of entire aerosol populations over realistic atmospheric timescales are possible. The underlying rationale in designing the experimental programme for the chamber facility outlined in this paper is to replicate aspects of atmospheric aerosol systems and, by close coordination of the experimental design with detailed model investigation of the relevant processes, aim to capture and understand the characteristics of the system determining the radiative implications. Such chamber experiments may not provide detailed mechanistic understanding of all processes contributing to the aerosol development (though extension of probing techniques is always possible), but those most pertinent to their atmospheric significance may be investigated in sufficient detail to test process model description. Throughout this proceedings issue a broad range of photochemical gaseous and aerosol chambers is described. The development of one of the most mature and largest photochemical chamber facilities is reported on http://pah.cert.ucr.edu/~carter/epacham/. It has been predominantly used to develop and validate the SAPRC-99 oxidation mechanism and is reported in Carter et al. (2003). It has not, however, been used extensively to investigate aerosol processes. One of the largest and most well appointed chambers used to investigate aerosol processes is the Caltech chamber (Cocker et al., 2001). It has been used to investigate the formation of aerosols from the photooxidation of diiodomethane (CH2I2) (Jimenez et al., 2003) and to develop and validate a model of secondary organic compound partitioning (Griffin et al., 2003) based on an extensive programme of aerosol formation experiments. Studies in the Kamens group facility at the University of North Carolina (Jang et al., 2003) have resulted in the discovery of condensed phase reactions leading to secondary organic aerosol formation and stabilisation. This concept has been further developed in the findings of Limbeck et al. (2003) showing acid catalysed aerosol formation from isoprene and Kalberer et al. (2004), where the condensed phase products after aerosol formation were shown to increase in molecular mass in the oxidation of 1,3,5-trimethylbenzene. The latter investigation was conducted in one of Europe’s largest indoor chamber facilities at the Paul Scherrer Institute in Zurich. There are other facilities in Europe where a range of aerosol investigations have been conducted. These range from the indoor chamber at Jülich and the EUPHORE chamber at Valencia to a range of smaller chambers. Most have some instrumentation for aerosol sizing and composition measurements, and a range of programmes, mostly limited to SOA formation studies in biogenically and anthropogenically

Field Measurement and Modelling Motivated Design of a Programme

51

influenced air masses (OSOA, EXACT, IALSI) have been conducted in these chambers. A mechanistic understanding of the formation of aerosols from photochemically formed gas phase precursors has been a priority objective in these studies, developing aerosol production potentials and yields in a range of systems, to provide an experimentally-constrained means by which secondary organic gas-to-particle conversion can be incorporated into useable models. However, many of the recent studies indicate the importance of secondary condensed phase reactions and complicated (yet not necessarily atmospherically representative) aerosol formation behaviour in the presence of high concentrations of mixed gas phase parent hydrocarbons. In addition, theories describing aerosol formation in chambers are not directly comparable to models including aerosol dynamics since the initial stage of nucleation is not considered, little molecular cluster and ultrafine aerosol number information is available and unrealistically high reactant concentrations are frequently used. For this reason, a primary difference between the studies outlined in this paper and the frequent SOA yield type experiments carried out in other chambers is that the experimental programme is designed to investigate aerosol evolution and growth, but will not consider formation. This avoids mechanistic interpretation difficulties of the formation process in terms of coupled chemistry and microphysics, but obviously provides no insight into nucleation and formation mechanisms. The experiment design thereby focuses only those aerosol process aspects directly accessible to description by relatively straightforward models. Approach to Chamber Design The design of a new aerosol chamber has followed from the range of atmospheric questions to be addressed in the experimental programme. Using climatic implications as both initial driver and final goal, the variety of investigations was broadly defined. In addition, the currently available aerosol modelling tools and those under development largely dictate the scope of the processes to be investigated. This paper therefore aims to describe the scientific questions, the modelling tools describing these questions, the proposed experimental protocol and the chamber design in response to the requirements. Range of Atmospheric Questions to be Addressed Avoiding investigations into aerosol formation processes, there are several distinct types of experiments which may be used to explore various aspects of atmospherically important systems including. x Non-reactive equilibration and microphysical evolution. Conventionally, the atmospheric aerosol has frequently been investigated assuming equilibrium is instantaneously attained between all semi-volatile components. This is conceptually adequate when considering systems which are not thought to be subject to kinetic limitations such as imposed by gaseous or aqueous diffusion or interfacial transport. The most straightforward systems would be equilibration of a population of small single involatile salt particles under varying

Figure 1. Predicted hygroscopic growth factor of NaCl / (NH4)2SO4 aerosol at 75% RH as a function of particle size in metres.

52

G. McFiggans

humidity and temperature environments. Equally accessible to description by available modelling techniques would be consideration of equilibration of a similar population of mixed involatile salt particles in a range of composition mixing states. Even in such simple systems, there are interesting predictions which may be investigated, such as that illustrated in Figure 1. Accounting for the effects of curvature on composition, calculations show that one might expect particles below a certain size to remain undeliquesced at higher relative humidities than those greater than this size (using the model of Topping et al. (2005a). More complex situations arise when semi-volatile components are present and the gaseous concentrations of these components will be important. Extension of the size distribution to include larger particles will introduce mass transfer limitation due to diffusion to the larger particles. Mixed semi-volatile component particles in such distributions may exhibit disequilibrium compositions. Introduction of organic components into the particles, either in externallymixed distributions or internally-mixed with inorganic components, may also introduce interesting phenomena. There is increasing evidence that organic components suppress efflorescence with reducing humidity (Marcolli et al., 2004). In addition, it is thought that some organic components take a significant time to attain equilibrium with ambient water vapour (e.g. up to 10 hours for glutaric acid (Peng et al., 2001)). Such equilibrium and kinetic effects will have significant atmospheric implications. In contrast to, for example, electrodynamic balance studies, large chambers provide a means of investigating such effects on entire aerosol populations. In addition, they offer longer residence times than flow systems (such as conditioners in TDMA studies) and hence the ability to better probe the extent of disequilibrium. Simple investigation of the changing equilibrium composition may be interpreted using thermodynamic models whilst the evolution of aerosol populations in chamber experiments over atmospheric timescales requires coupled microphysical and thermodynamic models. x Physical and chemical evolution of target aerosol system in isolation and externallymixed with a different seed background aerosol population with gas phase photochemistry. It may be appreciated that the introduction of gas phase photochemistry into systems such as described above will introduce a range of complications of atmospheric importance. Photochemical production or removal of semi-volatile components will perturb equilibria and hence change the evolution of the aerosol population. Production of reactive condensable species may initiate aqueous chemical reaction, further perturbing the equilibria. The aqueous chemistry can, in turn, effect the gas phase chemistry initiating complex multiphase cycles. Even in simple inorganic systems, such cycling may not be straightforward, e.g. halogen cycling through sea salt aerosol (Sander and Crutzen, 1996; Vogt

Field Measurement and Modelling Motivated Design of a Programme

53

et al., 1999; McFiggans et al., 2000; 2002). Figure 2 demonstrates the evolution of gas phase chlorine species resulting from such seasalt activation. Both gas phase and size-resolved aqueous phase composition measurements will be required. Cycling in organic systems will be hugely complex, but are likely to occur in the atmosphere. Large chambers provide sufficiently long residence times to establish such cycles and monitor the changes in aerosol and gas phase composition and physical properties on realistic atmospheric timescales. x Reactive development of target aerosol system in isolation and externally-mixed with a different seed background population and constant / varying gas phase photooxidant environment. It may be envisaged that the properties of atmospheric aerosol will be significantly affected by oxidation. For example, it is known that the oxygen to carbon ratio in fine particulate organic material increases with time moving away from an urban source region as does the hygroscopicity of the particles (McFiggans et al., 2005). Figure 3 shows the changing representative mass fragmentation pattern from an aerosol mass spectrometer deployed in urban and rural locations. This picture is common to all mid-latitude northern hemisphere findings. The more remote distributions are always dominated by a large contribution from CO2+ (m/z = 44) whereas the urban distribution Figure 3. A comparison of the measured mass shows a characteristic common fragmentation pattern and mass loading distribution of hydrocarbon fragmentation. The aerosol organics in different photochemical regimes as aerosol population evolves in a defined by the CO and O3 levels (HC, LO – high CO, changing oxidising environment low O3 and vice versa), (from Alfarra et al., 2004). and both surface and bulk oxidative processes may be involved in the observed changes. It is well established that both coagulation and condensation contributes to microphysical aerosol population evolution, but in addition to emerging field evidence for high molecular weight organic compounds contributing to the organic aerosol loading in a range of environments (O’Dowd et al., 2004; McFiggans et al., 2005), an increasing body of evidence for condensed phase macromolecule formation (e.g. Kalberer et al., 2004; Baltensperger et al., 2005) and surface oxidation (e.g. Donaldson et al., 2005) is being provided by chamber and laboratory studies respectively. This indicates that condensed phase chemical reaction may provide a significant, if not dominant, contribution to aerosol composition evolution. A simplification of the real

G. McFiggans

54

atmosphere as may be provided in chamber experiments is required to investigate the competitive effect of condensed phase oxidation, condensation and coagulation. In each of the three broad classifications of investigation above, significant uncertainties exist in the quantification of the effects of the respective processes on the ability of the aerosol population to affect climate forcing. The specific systems to be studied are described in the methodology section below. Available Modelling Tools Thermodynamics: A model of multicomponent inorganic and organic aerosol thermodynamics, ADDEM, has recently been reported (Topping et al., 2005 a, b). Accounting for the effects of particle curvature on composition, the equilibrium phase state of all involatile and semi-volatile components can be predicted at any given gaseous partial pressure of the components. For example, the model can be used to calculate the equilibrium size of multicomponent aerosol at a given relative humidity. Microphysics – aerosol and cloud: A variety of microphysical models accounting for condensational growth, evaporation and coagulation of aerosol particles in the sub- and supersaturated regimes, and cloud droplets in supersaturation, are available to investigate the evolution of condensed material under conditions of varying temperature and humidity. Coupled photochemistry and aerosol: Relatively simple schemes incorporating oddoxygen, odd-hydrogen and odd-nitrogen photochemistry and the oxidation of CH4, lumped hydrocarbons and reduced sulphur compounds have been coupled to the aerosol microphysical model described above. The coupled model may be driven by 2- or 4-stream radiative transfer algorithms to replicate the photon flux at any atmospheric level. A version of the model exists, extending that of McFiggans et al. (2000, 2002), describing the reactive cycling of halogens through seasalt aerosol particles. This version includes pH-dependent aqueous halogen chemistry. Mixed-phase cloud processes: A variety of mixed-phase and ice cloud models exist, describing the homogeneous and heterogeneous formation of water droplets and ice crystals. The implications of aerosol particles on mixed-phase clouds may be evaluated if their ice nucleating properties are known. Specialised coupled models: There are a number of detailed process models which can be used to address various coupled systems. One of particular interest to the chamber studies is the coupled model of aerosol formation and evolution in the photochemical iodine oxide system. Atmospheric Requirements Pressure: The chamber is designed to investigate boundary layer processes and will therefore operate at ambient atmospheric pressure. That the chamber does not require pressurisation allows the use of a flexible FEP Teflon bag. Teflon film is gas permeable, but with advantages that generally compensate for the drawbacks. It is broadly transparent to radiation at boundary layer wavelengths with well characterised transmission characteristics (Cocker et al., 2001). In addition, despite documented wall losses and re-emission of reactive

Field Measurement and Modelling Motivated Design of a Programme

55

species, Teflon is largely inert. Cleaning and flushing procedures will allow satisfactory operation in experiments involving predominantly unreactive aerosol systems. Humidity: The atmosphere is moist. The planetary boundary layer water vapour mixing ratio is of the order of a percent and relative humidity will range from low tens of percent to supersaturation. During the frequent cycling of air through the turbulent cloud-free or cloud-capped boundary layer, the atmospheric aerosol may experience relative humidities ranging from below the efflorescence point of pure salt solutions to supersaturations at which all but the smallest particles will activate into cloud droplets. A high purity water humidification system, humidity control and water vapour measurement are required in any aerosol chamber charging system. Light: Although many dark processes affect the evolution of aerosol populations (e.g. coagulation, condensation, evaporation, surface and condensed phase oxidation by ozone and NO3 radicals), a coupled atmospheric gas / aerosol system evolves in fluctuating solar illumination and will be subject to a range of photochemical processes. To investigate their effects, a chamber must be equipped with controllable illumination. One solution is to locate the chamber outside. Avoiding problems with naturally varying light intensity and exposure to inclement weather, indoor chambers must be provided with artificial lights. Artificial lights do not simulate natural sunlight across the full ground level solar spectrum, therefore photolysis rates between species will vary differently under natural and artificial irradiation. This presents various problems with characterisation and fluctuating intensity with wavelength investigated in detail elsewhere (Carter et al., 1995). Such issues must be accounted for when interpreting results. Artificial lights do have the advantage of allowing reproducible photolytic conditions between experiments. It has therefore been decided that the current chamber will use an artificial light source with the spectral characteristics monitored frequently, relating the investigated processes to atmospheric conditions using an explicit radiative transfer algorithm. Temperature: The lower limit of the range of temperatures to be probed within the chamber experiments is provided by the phase changes of water and the upper limit by reasonably expected boundary layer maximum. It is not proposed to investigate fog or cloud processes within the bag, so no operation below the dewpoint temperature at a given water vapour mixing ratio is envisaged. Neither is it envisaged that the chamber will operate above around 30qC. The lighting system will provide a large heat flux and the initial air charge will be warm due to heating in the high flow rate pumping system (see below). There will be no requirement for additional heating in illuminated experiments. There will, however, be a significant requirement for cooling. The temperature response of the system must be rapid and the thermal inertia will be high due to the large volume of air to be cooled. Aerosols in the Atmosphere: Atmospheric aerosol populations are complex. They are the result of variable heterogeneous primary emission sources, modification by superimposed effects of coagulation, condensation, evaporation and cloud processing, of various homogeneous and heterogeneous secondary formation processes and of size-dependent composition. They comprise a few inorganic salts, supposedly well-characterised in terms of their physico-chemical properties; a range of unmodified and reacted crustal mineral materials; elemental carbon / soot from combustion processes and hundreds, if not thousands, of largely unidentified primary and secondary organic compounds. This complexity in the composition of the particulate matter is only one dimension in the challenges provided by studies of atmospheric aerosol. The particles range in size from a few nanometres in diameter to many tens or even hundreds of microns. These four or five decades in diameter correspond to around 12 to 15 orders of magnitude difference in mass between the smallest and largest

G. McFiggans

56

particles. It can be seen that one or two large particles can vastly dominate the mass of the many thousands or tens of thousands of smaller particles. Unfortunately, however, the climate forcing resulting from aerosols is more strongly a function of the number of particles than the mass. For example, reflectivity and persistence of clouds is dependent on cloud droplet number which in turn depends strongly on the aerosol number. It is the composition of particles around the threshold activable dry size (somewhere near 100 nm diameter) which determines cloud droplet number, not the average aerosol bulk composition. A further significant complication is provided by the mixing-state in terms of the chemical composition and in terms of the derived physical properties. Each of the many components in an aerosol population may exist in a given size range of particles; the properties of the population will be significantly different if the components coexist in the same particles (“internally mixed”) or are in separate particles within the size range – “externally mixed”. Evidence from instruments measuring the water affinity of a particular size range indicate a range of affinities for a given size range in most locations (Cubison et al., 2005a). This indicates a range of compositions for any given size. Taken together, this evidence implies that extreme care must be taken when claiming that studied aerosol systems are of atmospheric relevance. Appropriate systems must be probed in an appropriate manner to make any inferences about their potential radiative impacts. Representative Gas and Aerosol Concentrations: It is usual for laboratory experiments and even many chamber experiments to be conducted at elevated concentrations of both gaseous components and aerosol concentrations. Unlike gas phase reactions which tend to be of relatively low order, aerosol microphysical process dependencies range over very many orders in concentration. For example, the nucleation process in the iodine oxide system is at least 4th order in atomic iodine concentration. Coagulation processes are dependent on particle number concentration and will be in competition with condensation and evaporation in their contribution to aerosol population evolution, both of which are dependent on something similar to the surface area (a surface area modified for a transition regime diffusion correction is one way of considering this) rather than the number concentration. The concentration regime in terms of gaseous components and in terms of several moments of the aerosol distribution must therefore all be considered when designing coupled system experiments. Undoubtedly the simplest means to achieve this conceptually is to conduct all experiments at close to ambient conditions. This may, however, present very many practical difficulties. Practical Considerations in Chamber Design Available Instrumentation: One of the most obvious constraints on chamber design and experimental programme is the availability of instrumentation to probe the quantities of interest. As outlined above, it is frequently necessary to simultaneously monitor aerosol physical and chemical properties in addition to gas phase chemical composition. A wide range of instruments are available for deployment in ground and aircraft-based field measurement within the Atmospheric Sciences Group at the University of Manchester. These will be available for deployment in the chamber in scheduled campaign mode and otherwise periodically when not in field use. In addition, there are a range of core instruments available for all experiments. The instrumentation suite is broadly classified as follows: Aerosol size distribution determination: A Differential Mobility Particle Sizer (DMPS) system comprising two (one short and one medium) Vienna-type Differential

Field Measurement and Modelling Motivated Design of a Programme

57

Mobility Analysers (DMAs, Winkelmayer et al., 1991) coupled to TSI model 3025A and 3010 Condensation Particle Counters (CPCs) with sheath to sample flow controlled volumetrically to 10:1 is continuously available to provide size distributions from 3 to 800 nm diameter (dry or at known relative humidity). The following sizing instruments are available periodically. A Grimm model 1.109 optical particle counter provides optical particle size distributions from 0.25 to 32 µm at ambient relative humidity by measurement of backscatter light intensity and inversion by Mie theory. A TSI model 3321 Aerodynamic Particle Sizer (APS) provides ambient aerodynamic particle size distributions from 0.5 to 20 µm diameter. A TSI model 3081 Scanning Mobility Particle Sizer (SMPS) is available for aerosol sizing across one of two standard sub-micron ranges. In addition, a variety of CPCs with differing size-cuts can provide rapid nano-particle fluctuation measurements. Using combinations of the above sizing techniques alongside vacuum aerodynamic diameter measurement from the Aerosol Mass Spectrometer (se below), it is possible to obtain information about particle morphology and density (DeCarlo et al., 2004). Aerosol composition, derived property and mixing state determination: An Aerodyne Aerosol Mass Spectrometer (AMS, Jayne et al., 2000) provides quantitative sub-micron nonrefractory aerosol component mass distributions by vacuum aerodynamic size. The loadings are commonly reported as sulphate, nitrate, ammonium and total organic distributions. In addition, the AMS has been used to discriminate between the characteristic organic mass fragmentation patterns in different locations (e.g. Figure 3). In laboratory and chamber experiments with aerosol of known initial composition, the AMS can provide useful mechanistic process information (Alfarra et al., 2005). A Hygroscopicity Tandem Differential Mobility Analyser (HTDMA; Rader and McMurry, 1986) comprising a medium length Vienna-type DMA coupled in series to a humidifier, a second DMA and a TSI model 3010 CPC is used to provide measurements of the affinity of the particles for water. The first DMA selects a pseudo-monodisperse dry cut from an ambient population, the humidifier subjects the cut to a known controlled relative humidity and the final DMA scans across an appropriate size range to determine the hygroscopic diameter growth factor, GFD; the amount by which the particles of this size grow in response to the changing humidity. The true ambient aerosol GFD is retrieved by inverting the observed distribution using an optimal estimation method (OEM, Cubison et al., 2005b). The HTDMA may also be used to infer the aerosol population component mixing state. The range of GFD at a particular size indicates the degree to which the compositions at this size vary between the particles. Combinations of AMS and HTDMA analyses will not yield unique solutions to the mixing state in the ambient atmosphere, but in chamber experiments with known initial composition, combined with some limited single particle composition information from the AMS, will provide almost all required composition information. Gas phase component concentrations: A number of standard gas analysers are routinely available for determination of gaseous mixing ratio of CH4, CO, O3, NO, NO2 and SO2 at moderate atmospheric concentrations. GC determination of a limited number of NMHC species is also available for appropriate experiments. An Aerodyne Quantum Cascade Tunable Diode Laser Absorption Spectrometer (QC-TDLAS) is periodically available for NH3 measurements using the line at 967 cm-1 and for NO2 measurement at 1606 cm-1. There are several potential experiments which may benefit from measurement of other trace species. The Differential Optical Absorption Spectrometer (DOAS) described below may be used for determination of IO, OIO, BrO, NO3, HCHO, SO2 and other trace species. Probing the potential for aerosol processes to affect impact on radiation: A key feature and major objective of the chamber facility is to investigate the changing potential of the aerosol populations to affect radiative forcing as they evolve chemically and physically

58

G. McFiggans

and thus to improve quantification of the effects of competing aerosol processes on climate. To fully quantify all potential effects it is necessary to investigate direct scattering of short and longwave radiation and longwave absorption in addition to indirect forcing by warm cloud activation of cloud condensation nuclei (CCN) and heterogeneous nucleation of ice nuclei (IN) in mixed phase / cold clouds. A Droplet Measurement Technologies CCN spectrometer, capable of measuring at supersaturations as low as 0.07 % is continuously available. An ice nucleus (IN) counter, built to the design of Rogers et al. (2001), is available to monitor the ability of the aerosol to act as ice nuclei. A DOAS system, comprising a Xenon arc lamp, active White cell optics traversing the minor axis of chamber coupled by optical fibre to a Princeton spectrometer with CCD detector is under construction. The DOAS system is not able to fully quantify direct aerosol scattering and absorption but will provide aerosol optical depth which may be interpreted alongside the direct size distribution measurements. Figure 4 shows a schematic of the chamber layout. It can be seen that the instrumentation requiring sample extraction is split between local and remote. The local instrumentation is located in the chamber laboratory, either fully enclosed in the environmentally controlled area (such as DMPS and HTDMA) or nearby. These instruments are on short bespoke inlets to minimise losses of, for example, ultrafine particles or “sticky” gases such as NH3. The “remote” instrumentation is located in the main aerosol laboratory above the chamber and allows the interfacing of the chamber to those instruments which are normally deployed in field experiments. Characteristic Requirements Common to All Chambers: The control of the environment will limit the useful experiments which are possible within the chamber. The environmental control provides the primary infrastructural requirements for the experimental programme. Charge supply: gas cleaning and aerosol generation: In experiments concerned primarily with investigation of nonreactive aerosol equilibration and microphysical processes, the requirements for gas scrubbing are less rigorous than for photochemical process studies and investigation of aerosol formation from parent hydrocarbons. However, a generally clean air supply is still required, particularly

Field Measurement and Modelling Motivated Design of a Programme

59

for those experiments investigating condensed phase oxidation processes. Particle filtration will be critical, as will the aerosol generation. The flush – fill cycling and experimental charge preparation system is driven by an oil-free 3 kW rotary side-channel blower. This is connected via 50 mm I stainless steel tubing via a series of pneumatically-operated 2- and 3-way fullbore ball valves and a series of filters. Currently gas scrubbing incorporates high flow rate activated carbon (Sofnocarb) and ultrafine particle (HEPA) filtration. Three additional conditioners will be added to the clean air supply (Molecular sieve, Moleculite and Sofnolime) to dry, convert CO to CO2 and remove CO2. The pump and all valves are computer controlled to cycle through chamber filling and flushing at greater than 1 m3 min -1 to enable many cleaning cycles between experiments. A corona discharge ozone generator is provided in-line to scrub the chamber of reactive hydrocarbons during the cleaning cycles. The aerosol generation is provided by a number of laboratory spray and nebulisation rigs feeding a 200 litre stainless steel reservoir through any combination of four inlet ports. This allows generation of any combination of internal and external aerosol component mixture to be fed to the experimental charge. Trace gas doping of the chamber is also possible through the reservoir. The flow is switched through the reservoir after the last cleaning cycle when delivering the experimental charge. Illumination and light flux measurements: The chamber illumination is still under design. The entire chamber is housed within an extruded aluminium frame. This will be enclosed within an outer walled frame to which the lamps will be fitted. The primary consideration is that a representative atmospheric light intensity of an appropriate spectrum is provided. The means of delivery is currently unresolved but will either comprise xenon or argon arc lamps or a combination of blacklights and other strip lighting. Both have advantages and the reader is referred elsewhere for a discussion of their relative merits (Carter et al., 1995). In any case, it is necessary to determine the wavelength-resolved light transmission of the chamber film. In addition, preliminary characterisation of the actinic flux across the chamber will be carried out using a Bentham DTM 300 scanning spectroradiometer (most appropriate in the UV) and a diode array spectrometer (with broader wavelength range and capable of more rapid measurement). Both spectrometers are available and for characterisation, the chamber is equipped with an adapter for the inlet manifold, rails to traverse the chamber and a miniature 2S input optical assembly. Relative humidity and temperature control and measurement: The strong temperature and RH dependence of the properties of interest in the target aerosol systems require that these quantities are continuously monitored and controlled such that fluctuations within the chamber throughout the duration of each experiment do not introduce undue errors. A chilled mirror hygrometer reference is linked to a number of capacitive RH and T sensors to continuously monitor the chamber conditions at a number of locations. The water vapour content of the chamber requires a pure water supply linked to the charging facility through a commercial steam generator. The temperature (and hence RH) control system is under design in conjunction with the illumination. The temperature control is necessitated largely by the lamp cooling requirement and also by the great sensitivity of the partitioning of semi-volatile components to temperature fluctuations. The air temperature between the clad outer frame and the film will be controllable to between 10qC and 30qC with appropriate response times and to an acceptable tolerance, nominally around 1qC fluctuation immediately following the lights being turned on and 0.5qC temperature drift throughout the experiments. This requires significant air handling and heat exchange (of the order of 30 to 40 kW with 8 to 10 kW of blacklight illumination). The temperature control and water vapour supply provides the RH control.

G. McFiggans

60

Chamber Dimensions and geometry: A key consideration in chamber design is the minimisation of reactant / aerosol losses and hence the maximisation of experimental duration. A larger chamber provides smaller wall area per unit volume and therefore a reduction in the wall effects. The geometry minimising the surface area per volume is a sphere. For practical purposes and cost effectiveness a nearly cubic chamber of 18 m3 has been chosen. The volume of air in the chamber will reduce through an experiment due to the sample volume requirement of the instruments. The chamber is mounted on three pairs of rectangular extruded Figure 5. Current chamber infrastructure. aluminium frames. The central pair are fixed whilst the upper and lower pairs are counterweighted and are designed to move in response to the reducing volume throughout each experiment (and during each flush / fill cycle). The chamber expansion and contraction is synchronised to the valve and pump operation by computer control. Figure 5 shows the current status of the chamber. Experimental Programme A number of preliminary studies have been defined to explore the suitability of the chamber for the three basic types of investigation detailed previously. The studies have been designed to be conducted throughout the early evolution of the working chamber. Experiment 1 is predominantly driven by predictions from a single and mixed salt thermodynamic model and will overlap significantly with the characterisation of the chamber. Predictions of the hygroscopicity of single-salt inorganic aerosols suggest that their deliquescence RH increases with decreasing size due to the Kelvin effect. The implication is that smaller aerosols may remain dry at the same RH that larger aerosols of the same composition will be deliquesced. Whilst possible to probe this effect systematically using a HTDMA alone, the process would be laborious and error-prone; likewise for electrodynamic balance experiments. This experiment is arguably the most straightforward, but may have profound implications on the direct scattering of light and on the surface reactivity of, for example, marine aerosols. After the initial characterisation stage, when wall losses of single salt aerosols will be investigated over a range of constant humidities, experiments will be conducted by nebulising broad aerosol distributions from single salt solutions into the chamber below the efflorescence RH of the aerosols at the inlet temperature. The chamber temperature will then be lowered, the size and composition of the distribution monitored and extinction measured by DOAS as the RH increases. Predictions of the changes in the distribution will be used to examine whether the measured aerosol extinction changes are consistent with suppressed deliquescence with size. This experiment will ascertain whether it is necessary to capture solid precipitation and deliquescence variation as a function of aerosol size in numerical descriptions of aerosol in order to realistically describe their

Field Measurement and Modelling Motivated Design of a Programme

61

direct radiative importance. For single involatile salts there is expected to be no change in indirect scattering, since this will depend on activating particles on increasing RH above ice or water saturation. Experiment 2 extends experiment 1 to consider multiple salt aerosols and mixed inorganic / organic aerosols. It is uncertain how distributions of even simple multiple salt aerosols behave with varying humidity. Similar predictions for multiple salts using the model from Topping et al. (2005a) indicate complex precipitation of double salts and multiple hydrates leading to staged deliquescence which varies as a function of size. Only by coupling the multicomponent activity model to a consideration of multicomponent surface tensions and densities can such predictions be made. Repeating experiment 1 with a range of multicomponent salts will allow these predictions to be tested. The multicomponent inorganic model has been extended to consider organic components using a group contribution theory approach (UNIFAC) (Topping et al., 2005b). A series of experiments investigating broad distributions of mixed inorganic / organic aerosols will be used to investigate the affect of the organic components on efflorescence / deliquescence as a function of size. As in experiment 1, if the model predictions are confirmed, it will indicate that size dependence of the deliquescence of multiple salt and mixed inorganic / organic aerosols must be captured to predict their scattering. In addition, recent electrodynamic balance work suggests that efflorescence is suppressed with an increasing number of organic species (Marcolli et al., 2004). This will be directly tested in this experiment by generation of appropriate test aerosols. A further complication may arise in these test systems in that equilibration timescales for some organic species in changing humidity environments may be of the order of several hours (Peng et al., 2001). This cannot be tested in the limited residence time allowed in HTDMA systems and requires the extended residence time allowed by chambers such as the one proposed. A final potentially important possibility to be tested is how the effect on the aerosol to act as CCN is affected by differential evaporation of volatile components as a function of size due to the extended equilibration timescales in the presence of organics combined with differential deliquescence with size. It is the aim to compare the experimental results with forthcoming developments in the mixed inorganic / organic thermodynamic model. Experiment 3 is driven by observations of changing aerosol properties as air moves downwind of a pollution source. The experiment will investigate processes contributing to the ageing of aerosols in a polluted plume. It is widely recognised that the composition of aerosols in polluted plumes evolves as the plume mixes into the regional background. This change is associated with a modification to the aerosol physical properties and an increase in their affinity for water. Since their affinity for water determines the ability for the aerosols to directly scatter radiation and to activate into warm cloud droplets, it is important to investigate the processes leading to the composition changes in aerosols in a polluted plume. A systematic chamber study programme will be used to explore the likely processes leading to these changes, firstly in simple single component systems of proxies for atmospheric organic aerosols, then in systems using several representative organic compounds combined variously in external and internal mixtures with inorganic components. Therefore, unlike those investigated in many chamber experiments, the polluted aerosol proxies will not be generated by nucleation from parent hydrocarbon oxidation, but will be nebulised from solutions of representative condensed phase components measured in the atmosphere. The generated aerosols will be subjected to oxidising environments in the absence of significant concentrations of gaseous hydrocarbons. The aim of this experiment is to investigate whether either surface oxidation processes or condensed phase reactions contribute to the aerosol evolution, particularly in terms of changes in the hygroscopic properties, thus changing their

62

G. McFiggans

radiative forcing potential (detected by CCN spectrometer, IN counter and DOAS). Hygroscopicity and composition measurements (HTDMA and AMS) can be directly compared with predictions from the mixed inorganic / organic thermodynamic model described above. No complex gas phase hydrocarbon oxidation mechanism will be required, hence oxidant concentrations under these experiments will be reasonably easy to predict. It is the intention to parameterise aerosol property changes in terms of oxidant concentration for incorporation into a coupled photochemical microphysical model. This experiment contrasts with those previous chamber experiments investigating the evolution of SOA nucleated in the presence of parent hydrocarbons in that it uses aerosols generated from compounds representative of those found in ambient samples as the reactant aerosols. These compounds may not prove to be of comparable reactivity to ambient aerosols, but may provide alternative insights to reaction of aerosols generated in unrepresentative parent hydrocarbon concentrations. This will help move towards a model of thermodynamic multicomponent inorganic / organic aerosol evolution and a parameterisation for condensed phase oxidation in sectional aerosol model. Experiment 4 is driven by atmospheric observations and previous laboratory studies and will investigate the atmospheric implications of coastal new particles formed during daytime low tide. Exposure of Laminaria macroalgae to atmospheric and elevated ozone levels in the laboratory has demonstrated that this is the source of bursts of iodine-containing particles in coastal experiments (McFiggans et al., 2004). Whilst mechanistic uncertainties remain, the most important question concerns the atmospheric significance of the particle “burst” phenomenon. The macroalgal ozone exposure reactor will be used to generate high ultrafine aerosol concentrations, comparable to coastal bursts. These will be fed into the chamber charge and evolve for several hours in the gaseous environment from the reactor outflow, and also in the same gaseous environment but seeded with SO2 to replicate that available from marine DMS oxidation. The primary aim is to ascertain whether particles formed from macroalgal emissions grow to radiatively active sizes in atmospherically reasonable timescales. For comparison, particles generated from the photooxidation of molecular iodine and diiodomethane (CH2I2) will be investigated in the same manner. DOAS retrieval of molecular iodine, iodine monoxide, iodine dioxide and sulphur dioxide concentrations will ensure that concentrations are representative of the coastal environment and enable a model of iodine particle nucleation and growth (McFiggans, 2002) to be constrained and applied to the chamber system. This experiment will provide a measure of the potential for iodine particles emitted in the coastal environment to participate in radiative forcing. Current Status The previous section describes a medium and long-term experimental programme. The characterisation of the chamber is currently at an early stage. Preliminary characterisation of the air supply and aerosol generation system has been conducted and some simple experiments attempted to establish the ability of the chamber to retain aerosol. Table 1 shows the average lifetime of relatively unreactive, involatile NaCl particles as a function of size measured during several 3 or 4-day experiments. Figure 6 shows an interesting phenomenon illustrating the forthcoming difficulties with chamber studies at low, atmospherically representative aerosol concentrations. The chamber was charged with an involatile NaCl seed aerosol in otherwise particle free scrubbed air. The illumination was with ambient laboratory lighting and daylight from the laboratory window.

Field Measurement and Modelling Motivated Design of a Programme

63

The decay of the number concentration with time at all sizes is evident over the first 12 hours or so in the DMPS measured size distribution. After this there is an obvious onset of nucleation and significant growth followed by decay in the aerosol number. The cycles of particle loss and fresh nucleation is repeated over several days, with no introduction of precursor gases. The initial concentration of NaCl particles is comparable to a low atmospheric loading and the nucleation is at a fairly low level. This illustrates the background level of aerosol dynamics which is currently experienced. This will improve with increased scrubbing and cleaning procedures.

Table 1. NaCl aerosol lifetime as a function of measured dry size.

Figure 6. Aerosol size distribution evolution during the first loss characterisation experiment.

Conclusions A chamber facility has been designed to investigate a range of atmospheric aerosol processes driven by their potential to affect radiative forcing. A preliminary experimental programme has been defined to complement an ongoing model development framework. The chamber infrastructure is partially constructed and undergoing preliminary characterisation. Illumination, temperature and humidity control systems are at various stages of design and construction and the experiments will follow the phased development of the chamber as appropriate. Acknowledgements The author would like to thank the UK Natural Environmental Research Council for support under grant NE/C000579/1 and would like to give thanks to the organisers of the Zakopane workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, NATO: EST.ARW 980164

64

G. McFiggans

References Alfarra, M. R., H. Coe, J. D. Allan, K. N. Bower, H. Boudries, M. R. Canagaratna, J. L. Jimenez, J. T. Jayne, A. A. Garforth, S. Li, and D. R. Worsnop; Organics by location, Atmos. Environ. 38 (2004) 5745-57. Alfarra, M. R., D. Paulsen, M. Gysel, A. A. Garforth, J. Dommen, A. S. H. Prévôt, D. R. Worsnop, U. Baltensperger, H. Coe; A Mass Spectrometric Study Of Secondary Organic Aerosols Formed From The Photooxidation Of Anthropogenic And Biogenic Precursors In A Reaction Chamber. submitted to Environ Sci Technol. (Oct. 2004). Baltensperger, U., M. Kalberer, J. Dommen, D. Paulsen, M. R. Alfarra, H. Coe, R. Fisseha, A. Gascho, M. Gysel, S. Nyeki, M. Sax, M. Steinbacher, A. S. H. Prevot, S. Sjögren, E. Weingartner and R. Zenobi; Secondary organic aerosols from anthropogenic and biogenic precursors, Faraday Disc. 130 (2005) doi: 10.1039/b417367h. Carter, W. P. L., D. Luo, I. L. Malkina, and J. A. Pierce, Environmental Chamber Studies of Atmospheric Reactivities of Volatile Organic Compounds. Effects of Varying Chamber and Light Source, Final report to National Renewable Energy Laboratory, Contract XZ- 2-12075, Coordinating Research Council, Inc., Project M-9, California Air Resources Board, Contract A032-0692, and South Coast Air Quality Management District, Contract C91323., March 26, (1995). Carter, W. P. L.; Evaluation of the SAPRC-99 Chemical Mechanism using new environmental chamber data, Presented at the Gordon Conference on Atmospheric Chemistry, Big Sky Resort, Montana, September 712, 2003 (available from: http://pah.cert.ucr.edu/~carter/bycarter.htm) Cocker, D. R. 3rd, R. C. Flagan, J. H. Seinfeld; State-of-the-art chamber facility for studying atmospheric aerosol chemistry, Environ Sci Technol. 35 (2001) 2594-2601. Cubison, M., M. Gysel, H. Coe and G. McFiggans; Are particles ever internally mixed? submitted to Geophys. Res. Letts. (2005) Cubison, M. and H. Coe; Use of an optimal estimation technique to retrieve ambient hygroscopicity distributions, submitted to J. Aerosol. Sci. (2005). Donaldson D. J., B. T. Mmereki, S. R. Chaudhuri, S. Handley and M. Oh; Uptake and reaction of atmospheric organic vapours on organic films, Faraday Disc. 130 (2005) doi: 10.1039/b418859d Griffin, R. J., D. Dabdub, J. H. Seinfeld; Development and initial evaluation of a dynamic species-resolved model for gas phase chemistry and size-resolved gas/particle partitioning associated with secondary organic aerosol formation, J. Geophys. Res. 110 (2005) doi:10.1029/2004JD005219 Jayne, J.T., D. C. Leard, X. F. Zhang, P. Davidovits, K. A. Smith, C. E. Kolb, and D. R. Worsnop; Development of an aerosol mass spectrometer for size and composition analysis of submicron particles, Aerosol Sci. Technol. 33 (2000) 49-70. Jang, M., N. M Czoschke, S. Lee and R. M. Kamens;. Heterogeneous Atmospheric Organic Aerosol Production by Inorganic Acid-Catalyzed Particle-Phase Reactions, Science 298 (2002) 814-817. Jimenez, J. L., R. Bahreini, D. R. Cocker, H. Zhuang, V. Varutbangkul, R. C. Flagan, J. H. Seinfeld, C. D. O'Dowd and T. Hoffmann; New Particle Formation from Photooxidation of Diiodomethane (CH2I2), J. Geophys. Res. 108, (2003) doi: 10.1029/2002JD002452. Kalberer, M., D. Paulsen, M. Sax, M. Steinbacher, J. Dommen, A. S. H. Prevot, R. Fisseha, E. Weingartner, V. Frankevich, R. Zenobi and U. Baltensperger; Identification of Polymers as Major Components of Atmospheric Organic Aerosols, Science 303 (2004) 1659-1662. Lewis, A. C., N. Carslaw, P. J. Marriott, R. M. Kinghorn, P. Morrison, A. L. Lee, K. D. Bartle, and M. J. Pilling, A larger pool of ozone-forming carbon compounds in urban areas, Nature, 405 (2000) 778-781. Limbeck A., M. Kulmala and H. Puxbaum; Secondary organic aerosol formation in the atmosphere via heterogeneous reaction of gaseous isoprene on acidic particles, Geophysical Research Letters 30 (2003) doi:10.1029/2003GL017738. Marcolli, C., B. P. Luo, and T. Peter; Mixing of the organic aerosol fractions: Liquids as the thermodynamically stable phases, J. Phys. Chem. A 108 (2004) 2216-2224. McFiggans, G., Plane, J. M. C., Allan, B. J., Carpenter, L. J., Coe, H. and O'Dowd, C; A modeling study of iodine chemistry in the marine boundary layer, J. Geophys. Res. 105 (2000) 14371-14385.

Field Measurement and Modelling Motivated Design of a Programme

65

McFiggans, G. Modelling the formation of aerosol due to iodine photochemistry – from the laboratory to the atmosphere?, Proceedings of the CERC 3 Young Chemists Workshop - Atmospheric Chemistry and Particulate Matter, Magleås, University of Copenhagen, Denmark, 3rd-6th June 2002 McFiggans, G., R. A. Cox, J. C. Mössinger, B. J. Allan and J. M. C. Plan; Roles of iodine and nitrogen in the release of active chlorine from marine aerosols, J. Geophys. Res. 107 (2002) doi:10.1029/2001JD000383. McFiggans, G., H. Coe, R. Burgess, J. Allan, M. Cubison, M. R. Alfarra, R. Saunders, A. Saiz-Lopez, J. M. C. Plane, D. Wevill, L. J. Carpenter, A. R. Rickard and P. S. Monks, Direct evidence for coastal iodine particles from Laminaria macroalgae – linkage to emissions of molecular iodine, Atmos. Chem. Phys. 4 (2004) 700-713. McFiggans, G., M. Alfarra, J. D. Allan, K. N. Bower, H. Coe, M. Cubison, D. O. Topping, P. I. Williams, S. Decesari, C. Facchini and S. Fuzzi; Simplification of the representation of the organic component of atmospheric particulates, Faraday Disc. 130 (2005) doi:10.1039/b419435g O'Dowd, C. D., M. C. Facchini, F. Cavalli, D. Ceburnis, M. Mircea, S. Decesari, S. Fuzzi, Y. J. Yoon, and J. P. Putaud; Biogenically-driven organic contribution to marine aerosol, Nature 431 (2004) doi:10.1038/nature02959. Peng, C. G., and C. K. Chan; The water cycles of water-soluble organic salts of atmospheric importance, Atmos. Environ. 35 (2001) 1183-1192. Rogers, D. C., DeMott, P. J., Kreidenweis, S. M., Chen, Y.; A Continuous-Flow Diffusion Chamber for Airborne Measurements of Ice Nuclei, J. Atmos. Oceanic Technol. 18 (2001) 725-741. Sander, R. and P. J. Crutzen; Model study indicating halogen activation and ozone destruction in polluted air masses transported to the sea, J. Geophys. Res. 101 (1996) 9121-9138. Rader, D. J. and P. H. McMurry; Application of the tandem differential mobility analyser for studies of droplet growth or evaporation, J. Aerosol Sci. 17 (1986) 771-787. Topping, D. O., G. McFiggans and H. Coe; A curved multi-component aerosol hygroscopicity model framework: 1 – Inorganics, Atmos. Chem. Phys. Discuss. 4 (2004) 8627-8676. Topping, D. O., G. McFiggans and H. Coe; A curved multi-component aerosol hygroscopicity model framework: 2 – Including organics, Atmos. Chem. Phys. Discuss. 4 (2004) 8677-8726. Vogt, R., R. Sander, R. von Glasow and P. J. Crutzen. Iodine chemistry and halogen activation in the marine boundary layer: A model study; J. Atmos. Chem. 32 (1999) 375-395. Winklmayr, W., G. P. Reischl, A. O. Lindner and A. Berner; A New Electromobility Spectrometer for the Measurement of Aerosol Size Distributions in the Size Range from 1 to 1000 nm, J. Aerosol. Sci. 22 (1991) 289-296. Winklmayr, W., G. P. Reischl, A. O. Lindner and A. Berner; A New Electromobility Spectrometer for the Measurement of Aerosol Size Distributions in the Size Range from 1 to 1000 nm, J. Aerosol. Sci. 22 (1991) 289-296.

Chamber Simulations of Cloud Chemistry: The AIDA Chamber Robert Wagner, Helmut Bunz, Claudia Linke, Ottmar Möhler, Karl-Heinz Naumann, Harald Saathoff, Martin Schnaiter, and Ulrich Schurath Forschungszentrum Karlsruhe, Institute of Meteorology and Climate Research (IMK-AAF), Karlsruhe, Germany Key Words: AIDA chamber, Ice nucleation, Polar stratospheric clouds, Soot aerosol, IR spectroscopy

Introduction Since its initial operation in 1997, the AIDA aerosol and cloud chamber of Forschungszentrum Karlsruhe (Aerosol Interactions and Dynamics in the Atmosphere) has been established as a unique experimental facility to study multi-phase processes over a wide range of atmospheric conditions. Research activities include heterogeneous chemistry on aerosols, hygroscopic and optical properties of complex aerosol particles, homogeneous freezing of supercooled solution droplets, heterogeneous freezing and cirrus cloud formation, as well as formation and characterisation of polar stratospheric cloud constituents.

Figure 1.

Schematic cross section of the AIDA facility. AIDA scientific instrumentation and aerosol generation devices are located on four separate floors surrounding the thermostated housing. 39 flanges with diameters from 0.01 to 1.1 m are located on the chamber walls, equipped with feed throughs for sensors, sampling tubes, and optical windows for various in situ spectroscopic measurements.

The AIDA chamber, a cylindrical aluminium vessel of 4 m diameter and 84.3 m3 volume, is located inside a large isolating container whose interior can be cooled to any temperature between ambient and 183 K (Figure 1). Temperature control is achieved by ventilating air around the aerosol vessel with upward directed flow rates of up to 30000 m3/h. Spatial and temporal temperature inhomogeneity is less than ± 0.5 K throughout the whole 67 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 67–82. © 2006 Springer. Printed in the Netherlands.

68

R. Wagner et al.

vessel. Cooling to 235 K is achieved by evaporation of a cooling liquid (R404A) inside two heat exchangers. For further cooling down to the minimum temperature of 183 K, liquid nitrogen is directly evaporated in the heat exchangers, leading to cooling rates of about 6 K/h at a liquid nitrogen consumption rate of approximately 0.6 m3/h. In addition to the broad accessible temperature range, the chamber can also be evacuated with two vacuum pumps from atmospheric pressure to a final pressure of about 0.01 hPa. This not only allows experiments to be performed at different pressure levels but also provides highly clean and aerosol free experimental conditions. Typically, background aerosol number concentrations are below 0.1 cm-3 after refilling the evacuated reaction vessel with particle free synthetic air. Further technical information about the thermostated housing, the vacuum pump, and gas supply system is available on the AIDA web page, http://imk-aida.fzk.de. In the following, a brief summary of major AIDA research activities is given, the advantages of carrying out these experiments in the AIDA chamber are highlighted, and recent journal publications ultilisating AIDA are referenced. Main components of the AIDA instrumentation will be described in the succeeding part of the article, including a comprehensive section about the application of FTIR spectroscopy to retrieve not only trace gas concentrations but also important aerosol parameters such as composition, phase, and size distribution. The remainder of this contribution will concentrate on ice nucleation experiments in the AIDA chamber. The operation of the AIDA vessel as a moderate expansion cloud chamber will be described and the data analysis of a typical ice nucleation experiment will be presented in detail. AIDA research activity since 1997 – an overview Early investigations with AIDA focussed on the heterogeneous reaction of ozone and other trace gases like NO2, HNO3, N2O5, NO3, and HO2 with soot aerosol (Kamm et al., 1999; Saathoff et al., 2001). The large volume of the AIDA chamber permitted experiments on time scales of several days, much longer than in previous laboratory studies. Therefore, it was possible to distinguish between reaction probabilities on typical tropospheric aerosol lifetimes and short-time saturation and passivation processes. In addition, the experiments were performed over a wide range of temperatures and relative humidities. Further studies placed emphasis on the detailed physical and chemical characterisation of airborne Diesel soot and spark generated soot particles, combining experimental and modelling approaches (Naumann, 2003; Saathoff et al., 2003a). Data analysis focussed e.g. on the optical properties of soot aerosol in the 230 – 1000 wavelength range (Schnaiter et al., 2003), as accurate optical data are urgently needed to assess the magnitude of the direct climatic impact of soot aerosol by absorption of solar radiation (Jacobson, 2001). The precise determination of absorption and scattering coefficients of pure soot particles prepared the ground for experiments dealing with the effect of the mixing state of soot with other aerosol particles on its optical properties (Saathoff et al., 2003b). Ageing processes like heterogeneous coagulation and condensation may change the mixing state of soot during its atmospheric lifetime of about one week, leading to soot particles coated with non-absorbing layers of organic and inorganic material. Theory predicts a strong increase in the specific absorption cross section upon coating, leading to enhancement factors of up to four, depending on layer thickness and core diameter (Fuller et al., 1999). To validate the model predictions, the optical properties of soot particles coated with different layers of organic material and sulphate were measured with the comprehensive optical instrumentation of the AIDA chamber. The organic coating was generated in situ by ozonolysis of D-pinene in the presence of soot aerosol. Due to the long residence time of the particles in the AIDA chamber, this coating procedure could be repeated

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

69

several times, thereby gradually increasing the thickness of the non-absorbing organic layer. Data analysis clearly revealed an enhancement of the specific absorption coefficient of the soot particles upon coating by a factor of 1.8 at the maximum, thus confirming the previous model calculations. The formation of secondary organic aerosol (SOA) from the oxidation of biogenic compounds, used in the previous study to generate the organic coating of the soot particles, was also the central subject of independent AIDA experiments. Especially the ozonolysis of D-pinene, which is thought to be one of the major sources of SOA in the atmosphere, has been the focus of many previous laboratory studies. One major uncertainty, however, still remained the influence of temperature on the SOA yield. Therefore, due to the outstanding temperature control of the AIDA chamber, the yield of SOA material from the ozonolysis of D-pinene and limonene was recently measured in the temperature range from 253 K to 313 K. Data analysis, resulting in aerosol yields as function of temperature and aerosol mass, is in progress. First AIDA investigations dealing with multi-phase processes in the polar stratosphere started in 1998. From the beginning, these experiments were characterised by a fruitful cooperation with Max-Planck-Institute for Nuclear Physics, Division of Atmospheric Physics, Heidelberg. The research group in Heidelberg had developed a balloon-borne aerosol composition mass spectrometer (ACMS), allowing quantitative chemical analysis of polar stratospheric cloud particles (Schreiner et al., 2003). A similar spectrometer was used for simulation experiments in the AIDA chamber (Knopf et al., 2001). Initial studies focussed on the composition analysis of supercooled H2SO4/H2O solution droplets, acting as precursors for PSC particles (Zink et al., 2002). After demonstrating the suitability of the AIDA instrumentation to analyse composition, size, and mass concentration of the solution droplets in the temperature range from 235 K to 185 K, subsequent experiments dealt with particle growth and homogeneous freezing of the H2SO4/H2O droplets under simulated lee-wave conditions (Möhler et al., 2003). Particle freezing was induced by adiabatic expansion experiments in the AIDA chamber, which will be described in detail later in this contribution. In addition to aerosol composition mass spectrometry, satellite observations in the mid-infrared proved to be a valuable tool to derive chemical composition, phase, and size distribution of stratospheric sulphuric acid aerosols and polar stratospheric clouds. The retrieval of important aerosol parameters relies on accurate refractive index data sets for the various aerosol constituents, inferred from several recent laboratory measurements and comprehensively tabulated in the HITRAN database of spectroscopic parameters (Rothman et al., 2003). A comparison between individual low-temperature refractive index data sets, however, revealed significant spectral differences in important wavenumber regimes. These discrepancies motivated another set of AIDA experiments to test the accuracy of published refractive indices for supercooled sulphuric and nitric acid droplets. Aerosol composition and mass concentration of the solution droplets which were measured independently by various techniques could be directly compared to the retrieval results from the analysis of midinfrared extinction spectra, thereby assessing the accuracy of the available data sets of refractive indices (Wagner et al., 2003). Ongoing work focuses on the optical constants of supercooled ternary H2SO4/H2O/HNO3 solution (STS) droplets. STS droplets form at temperatures below 195 K during the polar night when nitric acid and water condense onto background sulphuric acid aerosols. This formation mechanism was successfully mimicked in an AIDA experiment by injecting supercooled H2SO4/H2O droplets (acting as background aerosol) at T = 194 K, followed by the addition of ozone and nitrogen dioxide. Gas phase

70

R. Wagner et al.

reaction between O3 and NO2 yielded N2O5, which then hydrolyzed in the liquid phase forming ternary H2SO4/H2O/HNO3 solution droplets with gradually increasing nitric acid concentration. This experiment will also be used to determine the uptake coefficient of N2O5 on liquid sulphuric acid and STS droplets at stratospheric temperatures. Most recent AIDA measurements in the framework of polar stratospheric cloud research addressed the nucleation of solid nitric acid dihydrate (NAD) from sub-micron (< 0.5 Pm) supercooled binary HNO3/H2O and ternary H2SO4/H2O/HNO3 solution droplets. These experiments will be analysed with the help of a process model to distinguish between volumeand surface-induced nucleation modes. First model runs indicate that the volume-based nucleation model compares much better with our experimental data (e.g. number concentration of the nucleated NAD crystals) than the surface-based scheme. This finding is important in the context of recent laboratory measurements on ice nucleation, revealing that homogeneous freezing of supercooled water droplets is a volume-dominated process at least for droplet radii well above 1 Pm (Duft and Leisner, 2004). However, experimental evidence that surface nucleation will also be unimportant for much smaller droplet radii typical for aerosol droplets in the atmosphere was still lacking. Another interesting aspect from the NAD nucleation experiments in the AIDA chamber became obvious from the analysis of the midinfrared spectra of the nucleated NAD crystals. The infrared spectral signatures indicated the formation of clearly aspherical, predominantly oblate NAD crystals with aspect ratios exceeding five (Wagner et al., 2004). Such a high degree of particle asphericity was never observed before in laboratory studies on nucleation of airborne nitric acid hydrate crystals, a difference which may be attributed to the longer observation time in the AIDA chamber. Obviously, nucleated NAD crystals gradually adopt a high degree of particle asphericity when slowly growing by mass transfer from unfrozen solution droplets in a supersaturated environment over a long period of time, a process also supposed to take place in the polar stratosphere. Studies on the homogeneous freezing of supercooled solution droplets and the heterogeneous ice nucleation on the surface of solid aerosol particles like soot or mineral dust turned out to be another promising area of application of the AIDA chamber. These processes are thought to play a crucial role for the formation and radiative properties of cirrus clouds. Data analysis concentrates on deriving freezing thresholds and critical ice supersaturations for different types of aerosols, thereby providing accurate input data for numerical models simulating cirrus formation (Haag et al., 2003). This demands great skill to precisely determine the freezing onset in the ice nucleation experiments. Due to the extended AIDA instrumentation, at least three independent methods can be relied upon to measure the onset of ice formation: Detection of > 0.5 Pm ice crystals with an optical particle counter, appearance of the characteristic extinction features of ice in the infrared spectra, and increase in the depolarization of back-scattered laser light. As indicated in the introduction, an exemplary ice nucleation experiment will be discussed in a separate section below. The first set of AIDA experiments in the context of ice nucleation focussed on the homogeneous freezing of supercooled sulphuric acid droplets in the temperature range from 235 K to 185 K (Möhler et al., 2003). Measured freezing threshold relative humidities agreed well with a recent activity-based parameterisation for describing homogeneous nucleation of solution droplets (Koop et al., 2000). According to this theorem, only the water activity of the aerosol particles but not their specific chemical composition determines the critical ice supersaturation for homogeneous freezing. By titrating the sulphuric acid solution droplets with ammonia in situ in the AIDA chamber, pure ammonium sulphate aerosol particles can be

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

71

produced and subsequently used to study their ice nucleation potential. Remarkably, preliminary results have revealed that homogeneous freezing threshold relative humidities for ammonium sulphate solution droplets do not fit the activity-based theorem, showing significantly lower critical ice supersaturations compared to supercooled sulphuric acid droplets (Mangold et al., 2004). At present, however, it cannot be excluded that in the course of the titration process of sulphuric acid by ammonia gas some partial crystallization has occurred, yielding solid ammonium sulphate or letovicite, (NH4)3H(SO4)2. These solids may then have acted as nuclei for the supercooled ammonium sulphate solution droplets in the freezing experiments, reducing the ice supersaturation threshold compared to homogeneous freezing. Future AIDA freezing experiments are being considered to clarify this issue, thereby being careful to avoid any crystallization of solids in the supercooled ammonium sulphate solution droplets. Further AIDA ice nucleation studies addressed the ice nucleation efficiency of pure soot particles as well as soot particles coated with sulphuric acid (Möhler et al., 2004a). This work was part of a project investigating the impact of particles from aircraft engines on contrails, cirrus clouds, and climate. The results of these freezing experiments demonstrate that the mixing state of soot and sulphuric acid aerosol particles crucially affects the freezing threshold. Pure soot particles may act as efficient deposition ice nuclei in the temperature range from 235 K to 185 K, whereas coating with sulphuric acid significantly increases the critical ice saturation ratios, albeit still lying below the ice supersaturation thresholds for homogeneous freezing of pure sulphuric acid solution droplets. Very recent AIDA experiments have revealed the extraordinary properties of mineral dust aerosol particles in heterogeneous ice nucleation, acting as highly efficient deposition ice nuclei at any temperature between 195 K and 240 K (Möhler et al., 2004b).

T, p

Figure 2 .

Schematic view through the AIDA chamber, showing the most important devices for aerosol generation as well as trace gas and aerosol measurements. Note the mixing fan at the bottom of the vessel for achieving homogeneous temperature and humidity conditions inside the AIDA chamber.

R. Wagner et al.

72 AIDA instrumentation

Major components of the instrumentation of the AIDA chamber are shown in Figure 2. For a detailed listing of the complete technical equipment for trace gas and aerosol measurements, the reader is referred once again to the AIDA web page, http://imk-aida.fzk.de. Aerosol generation and characterisation. A broad variety of aerosol generation techniques has been designed for producing fundamental aerosol types, including sulphuric acid, nitric acid, and ammonium sulphate solutions droplets, airborne soot and mineral dust particles, as well as internally mixed aerosol particles like soot coated with sulphuric acid. Here, we want to draw attention to some special aerosol generation devices, i.e. our instrumentation for producing soot particles of different morphology and chemical composition as well as the apparatus for coating soot particles with sulphuric acid. Laboratory studies on airborne soot particles frequently employ an artificial soot sample from a commercial spark discharge generator (GfG1000) as surrogate for combustion soot. In the AIDA chamber, the properties of soot particles emitted from a VW Diesel engine have also been investigated, being available as part of the technical infrastructure of Research Centre Karlsruhe. It became obvious that the optical properties of airborne Diesel soot particles differ markedly from those of the synthetic soot sample, revealing, e.g., an increase in the specific extinction cross section by a factor of two at visible wavelengths (Schnaiter et al., 2003). More recent AIDA experiments introduced a new flame soot generator, the so-called combustion soot aerosol standard (CAST) burner. By slightly adjusting the CAST operation conditions, i.e. variations in the air/fuel(propane) ratio, soot aerosol samples with gradually varying organic carbon content could be generated. First results indicate that differences in the chemical composition significantly alter optical properties as well as the ice nucleation potential of the soot particles (Möhler et al., 2004b). Internal mixtures of soot or other aerosols like mineral dust with sulphuric acid can be generated in a homemade coating apparatus. Synthetic air which is saturated with sulphuric acid vapour by passing through a heated H2SO4 reservoir is mixed with the soot aerosol in a heated mixing chamber. Condensation of sulphuric acid vapour on the soot particles is achieved by passing the mixture through a temperature gradient tube. Size distribution measurements taken at the outlet of the tube section confirmed that condensation of sulphuric acid only occurs on the soot particles, i.e. without the formation of new particles. Higher temperatures of the sulphuric acid reservoir lead to an increase in the thickness of the sulphuric acid coating layer. Size distributions of the aerosol particles in the AIDA chamber are measured with a differential mobility analyser (DMA, 0.03 – 0.8 Pm range), connected to a condensation particle counter. As part of the AIDA experiments are dealing with partially volatile supercooled solution droplets like H2SO4/H2O, the DMA system was modified for operation inside the cooled isolating housing of the chamber (Seifert et al., 2004). This allows for measurements at stratospheric temperatures and pressures (183 K to 298 K and 100 hPa to 1000 hPa), thereby avoiding evaporation of the aerosol particles. Number concentrations and optical diameters of ice crystals in the 0.5 Pm to 20 Pm range are measured with two optical particle counters (PCS2000 and WELAS, Palas), mounted below the aerosol vessel and using vertical sampling tubes to minimise sampling losses.

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

73

Humidity measurements Our reference instrument for measurements of water vapour in AIDA is a fast highprecision chilled-mirror hygrometer (MBW, model 373). For ice nucleation studies, however, special instrumentation is needed to detect the rapid changes in relative humidity during fast adiabatic expansion experiments in the AIDA chamber at very low temperatures and correspondingly low water vapour mixing ratios. A time resolution of 1 - 2 s is required to accurately determine threshold relative humidities of ice nucleation, which is much faster than the response time of the MBW under these extreme conditions. Thus, ice nucleation studies are performed in close cooperation with the ICG I group of Forschungszentrum Jülich, who contribute their Lyman-D hygrometer FISH which measures water vapour mixing ratios with a time resolution of about 1 s and a sensitivity in the order of 0.2 ppm (Zöger et al., 1999). Note that the FISH instrument is operated in an ex situ mode by sampling AIDA air through a heated inlet tube. As a consequence, the water mixing ratio measured by the FISH corresponds to the sum of interstitial (gas phase) and evaporated particle water. A valuable complement to recording the total water mixing ratio is the in situ measurement of the gas phase water concentration by tuneable diode laser (TDL) absorption. The diode laser which can be tuned in the range 1370 ± 2 nm is located in the laboratory area outside the cold AIDA containment. The laser beam is coupled via an optical fibre to an internal White-type multiple reflection cell, currently allowing an optical path length of about 130 m. This affords a sensitivity of better than 0.1 ppm water vapour at ambient pressure. Optical instrumentation Apart from the TDL system, a large number of additional in situ and ex situ spectroscopic techniques are used to study light absorption by trace gas molecules as well as light extinction (absorption + scattering) by aerosol particles in the AIDA chamber over a broad wavelength regime. For example, aerosol extinction can be measured without any frequency gap from 200 nm to 16.7 Pm using independent UV/VIS, NIR, and FTIR spectrometers. Here, we want to briefly introduce the set-up of two selected instruments: (i) the in situ laser scattering unit which allows for precise determination of the onset time of ice formation in AIDA expansion experiments, and (ii) the FTIR spectrometer - the analysis of aerosol extinction spectra in the infrared will then be demonstrated in the next section. (i) In this set-up, vertical polarized laser light from an Argon ion laser at 488 nm is directed horizontally through the AIDA chamber. Scattered intensities are detected in the forward (4° detection angle) and polarization resolved in the backward direction (176°), which enables to calculate the depolarization of the back-scattered laser light with a time resolution of 1 s. An increase in depolarization during AIDA expansion experiments sensitively indicates the formation of ice crystals, the non-zero depolarization being a result of their aspherical particle habits. (ii) FTIR spectra of trace gases and aerosol particles are measured in situ with a second White-type multiple reflection cell, yielding a horizontal optical path length of about 250 m. Spectra are recorded from 800 to 6000 wavenumbers at a resolution of 0.125 cm-1 for the analysis of gas kinetic experiments as well as 4 cm-1 when dealing with particle spectra, which proved to be sufficient to resolve their generally broad-band extinction features, as shown in the next section.

74

R. Wagner et al.

Retrieval of aerosol parameters from FTIR extinction spectra – a short survey The quantitative analysis of aerosol extinction spectra is based on the Lambert Beer equation (1). The measured optical depth W Q j at a specific wavenumber Q j may be expressed as

W Q j  log

I Q j I 0 Q j

L N ¦ n  Di V  Di ,Q j ln10 i 1

j 1" M , (1)

where L denotes the optical path length, n(Di) the number concentration of particles in a particular size bin Di of width 'D, and V(Di,Q j ) the size bin–averaged extinction cross section, given by

V  Di ,Q j

1 'D

Di 

'D 2

³

V  D,Q j dD. (2)

'D Di  2

As input for the retrieval algorithms, a M x N matrix of extinction cross sections, with N representing the total number of particle size bins and M the number of wavenumber points, has to be calculated (Arnott et al., 1997; Zasetsky et al., 2002). Extinction spectra of spherical solution droplets like H2SO4/H2O are computed with Mie theory, whereas the T-matrix method is frequently applied to account for the aspherical habits of e.g. water ice and nitric acid hydrate crystals, albeit being restricted to rotationally symmetric aspherical particles like spheroids and cylinders (Mishchenko and Travis, 1998). Literature data of the temperatureand frequency-dependent complex refractive indices of the various aerosol particles are required as input for both the Mie and T-matrix calculations. Using standard optimisation techniques, the summed squared residuals between experimental and calculated spectra are minimised using the individual size bin populations n(Di) as fitting parameters. In the case of H2SO4/H2O, HNO3/H2O, or H2SO4/H2O/HNO3 solution droplets, the fitting algorithm may be extended to not only retrieve size distribution parameters but also the aerosol composition, expressed in terms of wt% H2SO4 and/or wt% HNO3. Additionally, the aspect ratio of e.g. prolate and oblate spheroids or cylinders can be introduced as optimisation parameter to estimate the degree of particle asphericity when analysing infrared spectra of water ice or nitric acid hydrate crystals. To demonstrate the quantitative analysis of aerosol extinction spectra, Figure 3 provides a compilation of selected FTIR extinction spectra of various aerosol types, recorded during recent AIDA measurement campaigns (Wagner et al., 2003). The left figure shows measured infrared spectra of supercooled H2SO4/H2O solution droplets at 233 K and HNO3/H2O solution droplets at 192 K in comparison with best fit results from Mie calculations using the size distribution and the aerosol composition as adjustable parameters. The synthetic spectra rely on optical constants for H2SO4/H2O and HNO3/H2O solution droplets inferred from aerosol infrared spectra (Niedziela et al., 1999; Norman et al., 1999). Note the good agreement between measured and calculated droplet spectra, which is a first indication of the high accuracy of the refractive indices used in the Mie calculations. Even more impressive is the excellent agreement between the aerosol parameters retrieved from the FTIR extinction spectra and those independently measured with the additional AIDA instrumentation (e.g. filter sampling and humidity measurements). From the sulphuric acid droplet spectrum an aerosol composition of 55 ± 2 wt% H2SO4 and a sulphate mass concentration (derived from the retrieved size distribution and the density of the solution

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

75

droplets) of 500 ± 25 Pg/m3 was inferred. This compares well with the complementary analytical results, namely 56.0 ± 0.5 wt% H2SO4 and 470 ± 50 Pg/m3 sulphate mass concentration. An equally good agreement was observed for the HNO3/H2O system, pointing out how accurately important aerosol parameters can be retrieved from FTIR extinction spectra if precise refractive index data are available.

H2SO4/H2O, 232.8 K

0.10

measured Mie fit

0.08 0.06 0.04

optical depth

0.02

0.04 0.03

NAD spheroids, calculated (c) optical depth / a.u.

optical depth

0.12

HNO3/H2O, 191.9 K measured Mie fit

0.02

NAD spheres, calculated (b)

measured (a)

0.01 0.00 6000 5000 4000 3000 2000 1000 -1

wavenumber / cm

Figure 3 .

2000

1600 1200 -1 wavenumber / cm

800

Left side: Measured FTIR extinction spectra of supercooled H2SO4/H2O and HNO3/H2O solution droplets in comparison with best fit results from Mie calculations. Right side: Measured D-NAD extinction spectrum compared to calculated spectra using different assumptions about particle shape. See text for details.

The right panel of Figure 3 illustrates the influence of particle asphericity on band positions and intensities in the infrared spectrum of nitric acid dihydrate in the 2200 to 800 cm-1 regime. Spectrum (a) shows the infrared extinction of D-NAD crystals which were formed via homogeneous nucleation of supercooled HNO3/H2O solution droplets at a temperature of about 195 K in the AIDA chamber. The measured spectrum is compared with two synthetic D-NAD spectra; spectrum (b) was calculated using Mie theory for a particle diameter of 1 Pm, and spectrum (c) was calculated with the T-matrix code for randomly oriented oblate spheroids with an aspect ratio of 15 and an equal-volume sphere diameter of 1 Pm, using optical constants of D-NAD retrieved from thin-film measurements (Toon et al., 1994). It is obvious that our measured D-NAD spectrum much better compares with the Tmatrix calculation, thus indicating the formation of strongly aspherical oblate D-NAD crystals (Wagner et al., 2004). The application of FTIR extinction spectroscopy to study composition and phase changes during a typical AIDA ice nucleation experiment is shown in Figure 4. The left panel shows a series of FTIR extinction spectra recorded during an AIDA expansion experiment

R. Wagner et al.

76

studying the homogeneous freezing of H2SO4/H2O solution droplets at 210 K. As described in detail in the next section, freezing could be initiated by reducing the pressure in the AIDA vessel with a mechanical pump. This results in cooling rates of up to 4 K/min, whereby freezing of the supercooled sulphuric acid droplets occurs when a threshold relative humidity is exceeded. During the freezing experiment depicted in Figure 4, aerosol extinction spectra were recorded every 40 s with the FTIR spectrometer. The bottom spectrum shows the infrared extinction of the sulphuric acid aerosol at static temperature and pressure conditions before pumping. The broad extinction feature at about 3300 cm-1 is due to the OH stretching mode of liquid water, absorption bands of the sulphate and bisulphate ions are located between 1000 and 1500 cm-1. After expansion cooling is started, the increasing relative humidity first of all leads to a dilution of the H2SO4/H2O aerosol particles which absorb water vapour from the gas phase. This is nicely evidenced by the continuous intensity increase in the liquid water extinction band at 3300 cm-1. Later on, distinct spectral changes occur in the grey-coloured spectrum: the frequency shift of the OH stretching mode as well as the appearance of a new extinction band at about 800 cm-1 both clearly indicate the formation of ice crystals. This indicates that the threshold relative humidity where supercooled H2SO4/H2O solution droplets freeze was exceeded. Note that the two subsequent infrared spectra shown in Figure 4 (scaled by factors of 0.6 and 0.4) are then completely dominated by light absorption and scattering of the growing ice crystals.

0.3 optical depth

*0.6

c(H2O) / ppm

*0.4

60 55 50 45 40 35 30 25 20

c(H2O) / ppm

0.4

35 30 25 20 15 10 5 0

0.2 't = 40 s

0.1

0.0 6000 5000 4000 3000 2000 1000 -1

wavenumber / cm

Figure 4 .

FISH: Total water TDL: Water vapour

0

200

400

600

FISH minus TDL FTIR retrieval 0

200

400

600

time relative to onset of pumping / s

Left side: Monitoring the homogeneous freezing of supercooled H2SO4/H2O solution droplets with FTIR extinction spectroscopy. Right side: Comparison of independent water measurements in the AIDA chamber. See text for details.

Quantitative analysis of FTIR spectra series recorded during AIDA expansion cooling experiments allows for deriving time profiles of the number concentration of the ice crystals as well as the ice water content (IWC). This provides a unique possibility to validate the independent measurements of these quantities with the optical particle counter as well as the FISH and TDL instruments, as demonstrated on the right side of Figure 4. The upper panel

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

77

depicts time profiles of total (i.e. gas + condensed phase) and interstitial water vapour, measured by the FISH and by the TDL systems, respectively, during a homogeneous freezing experiment with H2SO4/H2O solution droplets at 224 K. Time zero on the relative time scale indicates the start of pumping. As the H2SO4/H2O mass concentration, and thus the water content of the supercooled liquid aerosol particles was quite low in this experiment, total water and gas phase water mixing ratios are identical until the freezing onset at about t = 60 s. After this time, the difference between total and the interstitial water directly yields the ice water content, as depicted in the lower panel of Figure 4 (right side). This time profile is compared with the ice water content retrieved from the simultaneously recorded FTIR extinction spectra. In this approach, the ice crystals were modelled as cylinders with a fixed aspect ratio of D/L = 0.7 (D diameter, L length of the cylinders). The ice water volume mixing ratios were calculated from the retrieved size distributions of the ice crystals. The excellent agreement between both independent methods to estimate the ice water content underlines the high precision and accuracy of the water measurements in the AIDA chamber, which provide an excellent basis for ice nucleation studies over a very wide temperature range (Mangold et al., 2004). The AIDA facility as moderate expansion cloud chamber for ice nucleation studies As already noted in the introductory sections, AIDA ice nucleation studies primarily focus on a precise determination of the critical ice saturation ratio for relevant aerosol types like H2SO4/H2O solution droplets, mineral dust and soot particles, as well as internally mixed aerosol particles. The ice saturation ratio Sice(T) pw  T (3) pw,ice T is defined as the ratio of the actual water vapour pw(T) and the saturation water vapour pressure over ice pw,ice(T) at the same temperature. Ice nucleation only occurs when a critical threshold relative humidity with respect to ice (RHi) is exceeded, which may be significantly larger than 100%. For example, 160-170% RHi must be exceeded to initiate homogeneous freezing of supercooled sulphuric acid solution droplets at temperatures below 210 K. On the other hand, mineral dust particles act as deposition ice nuclei already at critical ice saturation ratios well below 1.2. In the atmosphere, supersaturations with respect to the ice phase are induced by adiabatic expansion cooling of rising air parcels. In the AIDA chamber, ice saturation ratios of > 1.6 are established by controlled pumping, typically from 1000 to 800 hPa. The mechanical pump can be operated at variable pumping speeds to establish cooling rates which vary by more than an order of magnitude. Figure 5 depicts temperature and pressure profiles for two different pumping scenarios. Using the maximum pumping speed (from 1000 to 800 hPa in less than 4 min) leads to a cooling rate of initially more than 4 K min-1. At slow pumping speeds, cooling rates less than 0.1 K min-1 can be controlled. Note that the cooling rate in AIDA approximately obeys the Poisson equation (4) only at the very beginning of an expansion cooling experiment:

Sice T

T

§ p· T0 ¨ ¸ © p0 ¹

R Cp

(4)

Later on the cooling rate continuously levels off due to the increasing difference between the gas temperature and the wall temperature, which remains almost constant throughout the expansion. This gives rise to an increasing heat flux from the walls into the

R. Wagner et al.

78

chamber air and limits the maximum supersaturation with respect to ice which can be achieved.

Pressure (hPa)

1 000

950

1 900 850

2

800

Temperature (K)

240 238

1

236 234

4 K/min

2

232

0.1 K/min

230

0

10

5

15

Time (min) Figure 5 .

Time profiles of total pressure (top panel) and mean gas temperature (bottom panel) during two AIDA expansion cooling experiments with largely different pumping speeds.

Expansion cooling experiments are started after coating the walls of the AIDA chamber with a thin ice layer, i.e. at an ice saturation ratio Sice close to unity. The coating is made by filling the chamber with humidified air at some higher temperature and subsequent cooling to a lower temperature. Due to several internal heat sources (mixing fan, heated sampling tubes) the mean gas temperature is always a few tenths of a Kelvin higher than the wall temperature under static conditions, which explains why Sice is slightly less than unity before an expansion is started. As already indicated in the introductory section, very low background aerosol number concentrations of < 0.1 cm-1 can be established by evacuating the chamber to < 0.01 hPa and performing several flushing cycles with particle-free synthetic air. Afterwards, the desired seed aerosol is added to the chamber and characterised by e.g. filter sampling and size distribution measurements. Upon the start of expansion cooling, the calculated saturation water vapour pressure over ice pw,ice(T) in Eq. (3) decreases exponentially with decreasing AIDA gas temperature whereas the actual water vapour pressure pw decreases at most linearly with the total pressure during pumping. This leads to a rapid increase in Sice with a rate dSice/dt of typically 0.3-0.4/min. Note that pw is directly measured in situ with the tuneable diode laser instrument. The rate dSice/dt during expansion cooling is further enhanced by evaporation of ice from the chamber walls as a result of the almost constant wall temperature during pumping. The onset of ice nucleation after exceeding the freezing threshold relative humidity as well as time profiles of the number concentration of the growing ice crystals can be detected by various techniques. In the following, we will demonstrate this interplay of the various AIDA instruments, taking as example two ice

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

79

nucleation experiments with different flame soot aerosol probes from the CAST burner as ice nuclei (Figure 6). This comparison will also illustrate how sensitively the ice nucleation potential depends on the chemical composition of the seed aerosol (Möhler et al., 2004b). Using different air/fuel ratios, two distinctly different samples of CAST soot aerosol with 16% and 40% organic carbon (OC) content, respectively, were generated. The OC fraction was determined by thermographic analysis of the soot aerosols sampled on quartz filters, following a standard protocol as described in the literature (VDI 2465, Blatt 2). To study the influence of the OC fraction of the soot aerosol on its ice nucleation potential, two otherwise identical AIDA expansion experiments were performed using either the 16% OC or the 40% OC soot samples as seed aerosol. The pressure changes during pumping as well as the associated changes in gas temperature for the two experiments are shown in the upper panels of Figure 6 (solid lines: 16% OC, dotted lines: 40% OC). Note the excellent reproducibility of the profiles, whereby the ice nucleation properties of different aerosol types can be directly compared. For the same reason approximately the same soot aerosol number concentrations (about 1000 cm-3) were used in both experiments. Panel 3 shows the relative humidity with respect to ice (RHi) calculated using Eq. (3) from the measured gas temperature and water vapour concentration (circles: 16% OC, triangles: 40% OC). When the onset time of ice nucleation can be determined, the corresponding threshold relative humidity can be directly read off this graph. At first, we will focus on the expansion cooling experiment with the 16% OC soot sample. Here, three independent measurement techniques point to a sharp, well-defined ice nucleation onset after about 100 s, as indicated by the solid vertical line. The optical particle spectrometer (PCS2000) starts to detect a large number of particles larger than 0.5 Pm in diameter, providing clear evidence for the formation and growth of pristine ice crystals (panel 5). At the same time, the infrared spectral signatures also indicate the nucleation of ice particles. The ice crystal number concentrations retrieved from the FTIR spectra (shown as circles in panel 5) turn out to be in excellent agreement with the data from the optical particle counter. Furthermore, the growing ice crystals lead to an increasing back-scattered light intensity, as evidenced in panel 4. From panel 3, we can infer an ice saturation ratio of 1.45 at ice nucleation onset. After reaching a maximum value of about 1.50, Sice starts to decrease as a result of the increasing uptake of water vapour by the growing ice crystals due to their increasing surface area concentration. Ice particle measurements in the expansion experiment with 40% OC soot aerosol markedly differ from the 16% OC sample. Note that the optical particle spectrometer hardly detects any ice particles. Additionally, extinction signatures of ice are barely visible in the infrared spectra and there is only a weak intensity increase of the back-scattered laser light in course of the expansion. The number concentration of ice crystals is less than 10 cm-3, thus < 1% of the seed aerosol particles act as deposition ice nuclei. In contrast to the 16% OC experiment, no precise critical ice saturation ratio can be specified for the 40% OC soot sample. RHi continues to increase to 190% because very little water vapour is lost on the small surface area of the scarce ice crystals. In summary, the comparison of the two expansion experiments provides first evidence that a higher fraction of organic carbon notably suppresses the ice nucleation potential of flame soot particles.

R. Wagner et al.

80

Figure 6 .

Measured time profiles of pressure, gas temperature, relative humidity with respect to ice, back-scattered laser light intensity, as well as ice particle number concentration for two expansion cooling experiments with different flame soot aerosol samples from the CAST burner as seed aerosol (Möhler et al., 2004b). See text for details.

Concluding remarks Due to the availability of several excellent diagnostic techniques for comprehensive studies of the formation and growth of ice crystals, as exemplified in Figure 6, ice nucleation on various types of aerosol particles will continue to be a focus of AIDA research activities. Future studies are planned to systemically analyse the impact of the chemical composition of soot on its ice nucleation properties. Further experiments will address the impact of coatings on the ice nucleating efficiency of mineral dust particles as well as the role of environmental ice nucleating bacteria. The above diagnostics will soon be supplemented by a new optical device which is presently under construction in our laboratory: it will be designed to measure

Chamber Simulations of Cloud Chemistry: The AIDA Chamber

81

scattering phase functions of ice crystals which are simultaneously imaged by a CCD camera with microscope optics at 2 µm optical resolution. References Arnott, W.P., C. Schmitt, Y.G. Liu and J. Hallett; Droplet size spectra and water-vapor concentration of laboratory water clouds: Inversion of Fourier transform infrared (500-5000 cm-1) optical-depth measurement, Appl. Opt. 36 (1997) 5205-5216. Duft, D. and T. Leisner; Laboratory evidence for volume-dominated nucleation of ice in supercooled water microdroplets, Atmos. Chem. Phys. 4 (2004) 1997-2000. Fuller, K.A., W.C. Malm and S.M. Kreidenweis; Effects of mixing on extinction by carbonaceous particles, J. Geophys. Res. (Atmos.) 104 (1999) 15941-15954. Haag, W., B. Kärcher, S. Schaefers, O. Stetzer, O. Möhler, U. Schurath, M. Krämer and C. Schiller; Numerical simulations of homogeneous freezing processes in the aerosol chamber AIDA, Atmos. Chem. Phys. 3 (2003) 195-210. Jacobson, M.Z.; Strong radiative heating due to the mixing state of black carbon in atmospheric aerosols, Nature 409 (2001) 695-697. Kamm, S., O. Möhler, K.H. Naumann, H. Saathoff and U. Schurath; The heterogeneous reaction of ozone with soot aerosol, Atmos. Env. 33 (1999) 4651-4661. Knopf, D.A., P. Zink, J. Schreiner and K. Mauersberger; Calibration of an aerosol composition mass spectrometer with sulfuric acid water aerosol, Aerosol. Sci. Tech. 35 (2001) 924-928. Koop, T., B.P. Luo, A. Tsias and T. Peter; Water activity as the determinant for homogeneous ice nucleation in aqueous solutions, Nature 406 (2000) 611-614. Mangold, A., R. Wagner, H. Saathoff, U. Schurath, C. Giesemann, V. Ebert, M. Krämer and O. Möhler; Experimental investigation of ice nucleation by different types of aerosols in the aerosol chamber AIDA: implications to microphysics of cirrus clouds, Meteorol. Z. (2004) accepted for publication. Mishchenko, M.I. and L.D. Travis; Capabilities and limitations of a current FORTRAN implementation of the Tmatrix method for randomly oriented, rotationally symmetric scatterers, J. Quant. Spectrosc. Radiat. Transfer 60 (1998) 309-324. Möhler, O., O. Stetzer, S. Schaefers, C. Linke, M. Schnaiter, R. Tiede, H. Saathoff, M. Krämer, A. Mangold, P. Budz, P. Zink, J. Schreiner, K. Mauersberger, W. Haag, B. Kärcher and U. Schurath; Experimental investigation of homogeneous freezing of sulphuric acid particles in the aerosol chamber AIDA, Atmos. Chem. Phys. 3 (2003) 211-223. Möhler, O., S. Büttner, C. Linke, M. Schnaiter, H. Saathoff, O. Stetzer, R. Wagner, M. Krämer, A. Mangold, V. Ebert and U. Schurath; Effect of Sulphuric Acid Coating on Heterogeneous Ice Nucleation by Soot Aerosol Particles, J. Geophys. Res. (Atmos.) (2004a) submitted. Möhler, O., C. Linke, H. Saathoff, M. Schnaiter, R. Wagner and U. Schurath; Ice nucleation on flame soot aerosol of different organic carbon content, Meteorol. Z. (2004b) accepted for publication. Naumann, K.H.; COSIMA - a computer program simulating the dynamics of fractal aerosols, J. Aerosol Sci. 34 (2003) 1371-1397. Niedziela, R.F., M.L. Norman, C.L. DeForest, R.E. Miller and D.R. Worsnop; A temperature- and compositiondependent study of H2SO4 aerosol optical constants using Fourier transform and tunable diode laser infrared spectroscopy, J. Phys. Chem. A 103 (1999) 8030-8040. Norman, M.L., J. Qian, R.E. Miller and D.R. Worsnop; Infrared complex refractive indices of supercooled liquid HNO3/H2O aerosols, J. Geophys. Res. (Atmos.) 104 (1999) 30571-30584. Rothman, L.S., A. Barbe, D.C. Benner, L.R. Brown, C. Camy-Peyret, M.R. Carleer, K. Chance, C. Clerbaux, V. Dana, V.M. Devi, A. Fayt, J.M. Flaud, R.R. Gamache, A. Goldman, D. Jacquemart, K.W. Jucks, W.J. Lafferty, J.Y. Mandin, S.T. Massie, V. Nemtchinov, D.A. Newnham, A. Perrin, C.P. Rinsland, J. Schroeder, K.M. Smith, M.A.H. Smith, K. Tang, R.A. Toth, J. Vander Auwera, P. Varanasi and K. Yoshino; The HITRAN molecular spectroscopic database: edition of 2000 including updates through 2001, J. Quant. Spectrosc. Radiat. Transfer 82 (2003) 5-44.

82

R. Wagner et al.

Saathoff, H., K.H. Naumann, N. Riemer, S. Kamm, O. Möhler, U. Schurath, H. Vogel and B. Vogel; The loss of NO2, HNO3, NO3/N2O5, and HO2/HOONO2 on soot aerosol: A chamber and modeling study, Geophys. Res. Lett. 28 (2001) 1957-1960. Saathoff, H., O. Möhler, U. Schurath, S. Kamm, B. Dippel and D. Mihelcic; The AIDA soot aerosol characterisation campaign 1999, J. Aerosol Sci. 34 (2003a) 1277-1296. Saathoff, H., K.H. Naumann, M. Schnaiter, W. Schöck, O. Möhler, U. Schurath, E. Weingartner, M. Gysel and U. Baltensperger; Coating of soot and (NH4)2SO4 particles by ozonolysis products of alpha-pinene, J. Aerosol Sci. 34 (2003b) 1297-1321. Schnaiter, M., H. Horvath, O. Möhler, K.H. Naumann, H. Saathoff and O.W. Schöck; UV-VIS-NIR spectral optical properties of soot and soot-containing aerosols, J. Aerosol Sci. 34 (2003) 1421-1444. Schreiner, J., C. Voigt, C. Weisser, A. Kohlmann, K. Mauersberger, T. Deshler, C. Kröger, J. Rosen, N. Kjome, N. Larsen, A. Adriani, F. Cairo, G. Di Donfrancesco, J. Ovarlez, H. Ovarlez and A. Dörnbrack; Chemical, microphysical, and optical properties of polar stratospheric clouds, J. Geophys. Res. 108(D5), 8313 (2003) doi:10.1029/2001JD000825. Seifert, M., R. Tiede, M. Schnaiter, C. Linke, O. Möhler, U. Schurath and J. Ström; Operation and performance of a differential mobility particle sizer and a TSI 3010 condensation particle counter at stratospheric temperatures and pressures, J. Aerosol Sci. 35 (2004) 981-993. Toon, O.B., M.A. Tolbert, B.G. Koehler, A.M. Middlebrook and J. Jordan; Infrared optical constants of H2O ice, amorphous nitric acid solutions, and nitric acid hydrates, J. Geophys. Res. (Atmos.) 99 (1994) 2563125654. Wagner, R., A. Mangold, O. Möhler, H. Saathoff, M. Schnaiter and U. Schurath; A quantitative test of infrared optical constants for supercooled sulphuric and nitric acid droplet aerosols, Atmos. Chem. Phys. 3 (2003) 1147-1164. Wagner, R., O. Möhler, H. Saathoff, O. Stetzer and U. Schurath; Infrared spectrum of nitric acid dihydrate influence of particle shape, J. Phys. Chem. A (2004) submitted. Zasetsky, A.Y., J.J. Sloan, R. Escribano and D. Fernandez; A new method for the quantitative identification of the composition, size and density of stratospheric aerosols from high resolution IR satellite measurements, Geophys. Res. Lett. 29, 2071 (2002) doi:10.1029/2002GL015816. Zink, P., D.A. Knopf, J. Schreiner, K. Mauersberger, O. Möhler, H. Saathoff, M. Seifert, R. Tiede and U. Schurath; Cryo-chamber simulation of stratospheric H2SO4/H2O particles: Composition analysis and model comparison, Geophys. Res. Lett. 29 (2002) 46-49. Zöger, M., A. Afchine, N. Eicke, M.T. Gerhards, E. Klein, D.S. McKenna, U. Mörschel, U. Schmidt, V. Tan, F. Tuitjer, T. Woyke and C. Schiller; Fast in situ stratospheric hygrometers: A new family of balloonborne and airborne Lyman-D photofragment fluorescence hygrometers, J. Geophys. Res. (Atmos.) 104 (1999) 1807-1816.

New On-line Mass Spectrometer for Identification of Reaction Products in the Aqueous Phase: Application to the OH-oxidation of N-methylpyrrolidone under Atmospheric Conditions Laurent Poulain, Anne Monod, and Henri Wortham Laboratoire de Chimie et Environnement, CNRS FRE 2704, Université de Provence case 29, 3 place Victor Hugo, 13 331 Marseille cedex 03, France Key Words : Aqueous phase tropospheric simulation chamber, On-line mass spectrometry, N-methylpyrrolidone, Kinetic rate constants, OH-oxidation, Reaction products, Chemical mechanisms

Introduction The atmosphere is a complex medium where chemicals are released, dispersed by physical processes and oxidized by photochemical reactions initiated by solar radiation. One of the most efficient oxidants is the OH radical produced by complex photochemical processes (Atkinson, 2000). In the past, the studies of Volatile Organic Compounds (VOCs) were focused on the gas phase reactivity. In the 1980s, inter-relations between gas and aqueous phases in the troposphere started to be considered (Graedel and Weschler, 1981). It was recognized that the aqueous phase photochemistry of Water Soluble Organic Compounds (WSOCs) has an impact on the gas phase concentrations of key species such as OH, HO2 and O3 (Lelieveld and Crutzen, 1990; Monod and Carlier, 1999; Herrmann, 2003). More recently, the contribution of WSOC to aerosol hygroscopicity and their ability to act as cloud condensation nuclei was found (Gelencser et al., 2003; Claeys et al., 2004; Ervens et al., 2004). Cloud droplets, which contain numerous chemical species originating both from the gas phase and from the condensation nuclei, are a medium where complex chemistry and photochemistry occur. Several free radicals (e.g. OH, NO3, Cl, Cl2-, SO4-) are formed, which can react with dissolved organic matter. Among these radicals, one of the most efficient oxidizing species is the OH radical (Ervens, et al., 2003, Herrmann, 2003). N-methyl-pyrrolidone (NMP) is used in the industry as a solvent. Its atmospheric lifetime is moderate, 13 h, if [OH] = 106 molecules cm-3 (Aschmann and Atkinson, 1999). However, NMP is a highly soluble compound (KH = 6.4 104 M atm-1, Hine and Moorkerjee, 1975), thus it is likely to enter into tropospheric droplets. The present work was aimed at determining the OH-oxidation rate constant of NMP and the reaction products formed in the aqueous phase under tropospheric conditions. NMP is a medium sized molecule (Table 1), containing 5 carbon atoms, so its OH-oxidation may give rise to a large number of products. This study presents a new on-line technique suitable to identify as many reaction products as possible, during the reaction. Two different kinds of experiments were conducted to determine respectively the kinetics and the reaction products formed. Experimental Reagents. The chemicals used were: H2O2 (not stabilized, Fluka, 30% w); methyl ethyl ketone (MEK) (Fluka, more than 99.5%); methyl-iso-butylketone (MIBK) (Fluka, more than 99%); FeSO4 (Prolabo), acetonitrile (Acros, HPLC grade); methanol (Acros, HPLC grade); succinimide (Sigma Aldrich, 98%), methylamide (Sigma Aldrich, 40%); N-ethylacetamide (Sigma Aldrich, 99%); N-methylsuccinimide (Sigma Aldrich, 99%); N-hydroxymethylpyrrolidone (Sigma Aldrich); acetamide (Sigma Aldrich, 98%), formamide (Sigma 83 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 83–96. © 2006 Springer. Printed in the Netherlands.

L. Poulain et al.

84

Aldrich, 98%), N-ethylformamide (Sigma Aldrich, 99%), 2-pyrrolidone (Sigma Aldrich, 99%), dimethylacetamide (Sigma Aldrich, 99%). Solutions were prepared using purified water from a Millipore MilliQ system utilising the reverse osmosis, micro-filtration, nucleargrade deionization and activated carbon modules (the resistivity of the purified water was greater than 1.8u 107 : cm-1). Glass cleaning procedure. The work was performed with diluted solutions, in order to avoid any contamination. The reactor and all the sampling vessels were carefully cleaned according to the following procedure: 3 rinses with reverse osmosed water; 1 hour soaking in 2 % DECON detergent diluted with reverse osmosed water; 5 rinses with ultrapure MilliQ water; 2 hours soaking in 2 % HCl diluted with ultrapure MilliQ water; 5 rinses with ultrapure MilliQ water; 15 hours soaking in ultrapure MilliQ water; 3 rinses with ultrapure MilliQ water. Kinetic studies of OH-oxidation of NMP We have shown previously that OH radicals can be generated through the Fenton reaction (1) under specific experimental conditions (Monod, et al., 2005). This technique was used in the dark at pH = 2 for the kinetic study of the OH oxidation of NMP. Fe

2+

H+

+ H2O2

OH

+

Fe(OH)2+

(1)

The kinetic rate constant of the OH oxidation of NMP was determined using the relative kinetic method. The principle of this method is to measure the decay rate of the OHinduced oxidation of the reactant (NMP) relative to that of a reference compound (R) for which the OH oxidation rate constant is well known (reactions 2 and 3). In this study, two reference compounds were chosen: methylethylketone (MEK) and methyl-iso-butylketone (MIBK): NMP + OH

R

+

OH

kX

kR

products

(2)

products

(3)

where kNMP and kR are the rate constants of the OH oxidation of NMP and a reference compound, respectively. The corresponding kinetic equations can be written as follows: § >NMP @ 0 · ¸¸ ln¨¨ © >NMP @ t ¹

§ >R @ 0 · k NMP ¸¸ u ln¨¨ kR © >R @ t ¹

(4)

where [NMP]0, [R]0, [NMP]t, [R]t are respectively the concentrations of NMP and a reference § >NMP @ 0 · § >R @ 0 · ¸¸ versus ln¨¨ ¸¸ yields a linear curve with compound at times 0 and t. Plotting ln¨¨ © >NMP @ t ¹ © >R @ t ¹ a slope of kNMP / kR and an intercept of zero. To obtain kNMP r dkNMP, the median value and

New On-line Mass Spectrometer for Identification of Reaction Products

85

§ >NMP @ 0 · ¸ ln ¨¨ the standard deviation (r2V ) of the ratios © >NMP @ t ¸¹ were calculated and multiplied by kR. § >R @ 0 · ¸¸ ln ¨¨ © >R @ t ¹

To establish the Arrhenius expressions for NMP, the equations given in Monod, et al. (2005) were used for the reference compounds:

ln k MEK (T )

26.2 r 1.0  1600 r 300

(5)

ln k MIBK (T )

25.6 r 1.0  1200 r 300

(6)

T

T

The analyses of NMP were performed using a High Pressure Liquid Chromatograph (HPLC) equipped with a reversed phase C18 column (Alltima, Alltech) and a UV detector working at 212 nm. A binary methanol/water eluent was used (with a gradient from 20%/80% to 90%/10% in 10 min, then maintained so for 5 min), at 1 mL min-1. The analysis of MEK and MIBK was performed by derivatisation with 2,4-dinitrophenylhydrazine (2,4-DNPH), followed by HPLC analysis with a reversed phase C18 column (Alltima, Alltech) and a UV detector working at 360 nm. A tertiary eluent (CH3CN/H2O/CH3OH) was used (isocratic at 20%/40%/40% for 3 min, then a linear gradient to 5%/25%/70% in 10 min, then another linear gradient to 5%/15%/80% in 10 min, then maintained for 6 min). Each sample was prepared by removing an aliquot of 180 µL from the reactor and diluting it with 800 µL of 2,4-DNPH (0.36 mg L-1). These techniques produced sharp peaks in the chromatograms. Before each experiment, a calibration was carried out for each compound in a range covering the concentrations encountered during the experiments.

23

T (K) 276

22.6 22,6

298 308

ln (kOH)

23

322

22.2 22,2

k OH + NMP (10 9 M-1 s -1) relative to MEK relative to MIBK (triangles) (squares) 2.69 r 0.52 2.44 r 0.37 4.63 r 0.51 3.50 r 1.3 4.30 r 0.54

3.45 r 0.31 2.99 r 0.47 3.66 r 0.40

21.8 21,8

21,4 21.4 21 21 3.00 3,00

3.10 3,10

3.20 3,20

3.30 3,30

3.40 3,40

3.50 3,50

3.60 3,60

3.70 3,70

10- 3 / T (K-1 )

Figure 1.

Arrhenius plot and relative rate constants (inset) for the reaction of OH radical with NMP in the aqueous phase.

L. Poulain et al.

86

The kinetic studies were conducted at four temperatures covering the atmospheric conditions, from 276 to 322 K. The results are presented in Figure 1. Taking into account all the data presented, one obtains the Arrhenius expression (7):

( 24.4 r 1.1) 

ln(k OH )

(760 r 320) T

(7)

The results show that the aqueous phase OH oxidation of NMP is fast compared to that of other WSOC, and that temperature has little effect on the kinetics. Assuming that 15% of air contains clouds, and that [OHg] = 106 molec/cm3, [OHaq] = 10-13 M, the presence of clouds can reduce the atmospheric lifetime of NMP by 13 %, from 13 h to 11 h (for the estimation of global atmospheric lifetimes, see Monod et al. (2005). Thus, the aqueous phase OH oxidation has an important impact on the atmospheric behaviour of NMP. Because the kinetics of the reaction is fast, special attention has to be paid to the reaction products formed. Reaction products studies

To study the reaction products the OH oxidation of NMP in the aqueous phase, the continuous photolysis of H2O2 was used to produce OH radicals (reaction 8). H2 O2

+

hX

2 OH

(8)

The Fenton reaction (1) was not used here to avoid any artefacts due to reactions of FeII with any organic acids formed. The reaction took place in an aqueous-phase simulation chamber, which consisted of a 450 mL Pyrex vessel. Unbuffered aqueous solutions were continuously stirred and maintained at 298 r 0.2 K. The light source was a 300 W xenon arc lamp (see Monod, et al. (2005) for further details on the apparatus). The initial concentrations of NMP and H2O2 were 5×10-4 and 1×10-3 mol L-1, respectively. During the reaction, aliquots of the solution were sampled at regular time intervals with an automatic sampler, and stored at 0 °C for further analysis. HPLC analysis

The formation of N-methylsuccinimide (NMS), succinimide and 2-pyrrolidone (2P) (Table 1) was expected by Campbell and Striebig (1999) and Friesen et al. (1999) who performed OH-oxidation of NMP in alkaline solutions and by photocatalysis, respectively. We performed the HPLC analysis using the same C18 column as described above, and a UV detection at 202 nm. A binary eluent (80 %/20 % acetonitrile/water) was used at 1 mL min-1. The results showed that some of the expected reaction products were formed. Figure 2a shows the time profiles of NMP, NMS and succinimide, while Figure 2b shows the corresponding plots of product concentrations versus the concentration of NMP consumed. The latter plots clearly show that NMS was a primary reaction product, whereas succinimide was a secondary one. The carbon balance of NMS production is 55 ± 1 %. This indicates that other primary reaction products have to be formed. Figure 2 shows that one unidentified compound was a primary product. Furthermore, the HPLC chromatograms contained a lot of unidentified and unresolved peaks (Figure 3). Therefore, an on-line mass spectrometer was designed in order to identify more reaction products.

New On-line Mass Spectrometer for Identification of Reaction Products

5x10

-4

4x10

-4

3x10

-4

2x10

-4

1x10

-4

0 0 NMP

10000

20000

30000

40000

time (seconds) 2P Succinimide

NMS

50000

2x10

6

2x10

6

1x10

6

5x10

5

0 60000

unidentified compound

[NMS] and [Succinimide] (mol/L)

Fig 2 b

2x10

-4

2x10

-4

1x10

-4

5x10

-5

0 0

1x10 NMS

Figure 2.

-4

2x10

6

2x10

6

1x10

6

5x10

5

unidentified compound (peak areas)

concentration (mol/L)

Fig 2 a

unidentified compound (peak areas)

87

0 -4 -4 -4 -4 2x10 3x10 4x10 5x10 ' [NMP] (mol/L) succinimide unidentified compound

HPLC analysis: (a) time profiles of NMP, two identified reaction products, (NMS and succinimid) and one unidentified reaction product, during the OHoxidation of NMP in the aqueous phase; (b) concentrations of the products (or HPLC signal in arbitrary units, a. u.) as functions of the consumed NMP.

L. Poulain et al.

88

Figure 3.

Example of an HPLC chromatogram during the OH-oxidation of NMP in the aqueous phase at t = 5 h 46 min (20778 seconds).

Design of an on-line mass spectrometer To perform in situ analysis of the solution during the course of the reaction, the simulation chamber was directly connected to a mass spectrometer (MS). The solution was transported to the MS by an HPLC pump. The MS was equipped with an electrospray ionisation (ESI) unit and a triple quadrupole, thus allowing ESI-MS+, ESI-MS-, and ESI-MSMS+ analyses to be performed. The HPLC pump flow was adjusted to 0.1 mL min-1 in order to get a good compromise between maximum analytical sensitivity and minimum volume sampled from the simulation chamber. A 9 minutes delay between sampling from the chamber and the MS analysis was taken into account in the evaluation of results. Over 20 hours of a reaction run, the mass spectrometer was operated continuously, in 48 cycles. Each cycle was 25 minutes long and contained 3 successive modes: ESI-MS+, ESI-MS-MS+ and ESI-MS(Figure 4), in order to identify the time profiles of as many compounds as possible. Each mode was operated with a scan range of 20 – 1000 atomic mass units (amu). The stabilisation times mentioned in Figure 4 were taken into account in the analysis of results. t = 6 min

t=0

ESI-MS positive mode

t = 14 min

ESI-MS-MS positive mode

Argon pressure stabilisation in quadrupole 2

t = 25 min

ESI-MS negative mode

Pressure stabilisation in quadrupole 2

Figure 4. A 25 minute operation cycle of the mass spectrometer directly connected to the simulation chamber during the OH oxidation of NMP in the aqueous phase. The cycle was repeated 48 times, to cover the 20 hours of a reaction run.

New On-line Mass Spectrometer for Identification of Reaction Products

89

Results. The experimental protocol allowed us to obtain, for the first time (to our knowledge), the time profiles of a large number of reaction products (around 66 species). The ESI-MS+ scans revealed around 40 products formed, and the ESI-MS- scans around 26 products. Figure 5a (full scale) shows the decrease of the NMP concentration during the reaction, at m/z = 100 amu in the positive mode (henceforward noted as 100+ amu), while Figures 5a (zoom) and 5b (full scale and zoom) show the formation of several oxidation products. Some of the products were formed initially and then destroyed, certainly due to their reaction with OH radicals (for example m/z = 86+, 114+, 130+ and 115-). Up to 300 amu, in both modes, one can observe the formation of compounds those masses are regularly spaced by 'm/z) = 14 to 18 amu (Figure 5a and 5b). This can be the consequence of the increasing oxidation level of the NMP ring, and/or the consequence of the formation of hydrated reaction products. Although the scan was performed up to 1000 amu, no signal was observed above 300 amu. Excellent reproducibility of the mass spectra was observed in three experiments. Discussions on the mass spectra. The identification of the products was performed by comparison with standard solutions (when available) and using ESI-MS+, ESI-MS-, and/or ESI-MS-MS+ scans. The results are shown in Tables 1 and 2. NMP was detected at 100+ amu. Interestingly, its dimer was detected simultaneously at 200 amu. Tests with standard solutions of NMP showed that the dimer was formed in the electrospray unit of the mass spectrometer, and the time-profile of the MS signal at 200+ amu seems to confirm that this dimer was not formed in the photoreactor (Figures 5a and 6). +

The time-profile of the MS signal at 100+ amu (Figure 6) is different from the NMP signal obtained by HPLC (Figure 2). The HPLC shows an exponential decrease in NMP, whereas the MS signal at 100+ amu shows a significant shoulder between 8000 and 22000 seconds, which is also observed in ESI-MS-MS+ scans at 58 amu. The shoulder can be explained by formation of a product of the same mass as NMP. Succinimide produced during the reaction was identified at 100+ amu, and the ESI-MS-MS+ scans of standard solutions revealed the same daughter ions as for NMP, at 58 and 72 (major) amu (Table 1). Fortunately, tests with standard solutions of succinimide showed that this compound and its hydrate can be detected in ESI-MS- scans at m/z = 97- and m/z = 115- (Table 2), and thus can be distinguished from NMP (Figure 6). In the HPLC analysis, N-methylsuccinimide (NMS) was identified as one of the main primary reaction products (Figure 2). Tests with standard solutions showed that NMS is detected in ESI-MS+ scans at m/z = 114+, and its hydrated form is detected in ESI-MS- scans at m/z = 130-. However, an excessive carbon balance (§ 120 %) was obtained with the MS analysis at m/z = 130- (calibrated with standard solutions of NMS), in disagreement with the 55 ± 1 % obtained by HPLC. This can be due to the formation of another product of equal mass. 1-Formyl-2-pyrrolidone (FP) has exactly the same mass as NMS, and was previously identified as an important primary reaction product during the OH oxidation of NMP in the gas phase (Aschmann and Atkinson, 1999). The ESI-MS-MS+ scans of mass 114+ during the experiments show the formation of two daughter ions: one at m/z = 86, which corresponds to NMS, as verified with standard solutions, and one at 72. FP is not available commercially, so we could not perform separate tests for this compound. However, the major fragmentation pathway of FP should include the loss of •CH2-C•(O) and/or •CH2-CH2-CH2•, by analogy with NMP, succinimide and NMS. The fragmentation pathway of FP leads to a major daughter ion at m/z = 72+, thus confirming the formation of FP during the reaction. Furthermore, FP may

90

Figure 5.

L. Poulain et al.

Full scale and zoom 3D plots of the MS signal versus mass and time during the OH-oxidation of NMP in the aqueous phase (scan range 20 to 300 amu), (a) in the positive mode, and (b, facing page) in the negative mode. See Table 1 for the identification of the observed masses.

also be responsible for the unidentified peak in HPLC (Figure 3), which signal shows that it is a primary reaction product (Figure 2). It is known that hydroxyl compounds can be formed in the aqueous phase during OH oxidation processes (Von Sonntag and Schuchmann, 1997), so this pathway was explored. At m/z = 116+ amu, Figures 5a and 6 show the formation of an important reaction product, which can be 5-hydroxy-N-methylpyrrolidone (5-HNMP) and/or Nhydroxymethylpyrrolidone (NHMP) (Table 1). The formation of both compounds was confirmed by ESI-MS-MS+ scans, with two daughter ions produced and detected at m/z = 88 (5-HNMP) and at m/z = 73 (NHMP). This is in good agreement with the major fragmentation pathways of NMP, NMS and succinimide. Furthermore, this was confirmed by tests with standard solutions of NHMP, which also allowed us to observe its hydrated form at m/z = 132- (Table 2). The formation of 2-pyrrolidone (2P) was suggested by Campbell and Striebig (1999) and Friesen et al. (1999). We observed its formation at m/z = 86+ amu (Figures 5a and 6), which was confirmed by ESI-MS-MS+ scans (Table 1). The formation of smaller compounds such as methylamine, formamide, NMF, N-methyl-4-aminobutanoic acid (in ESI-MS+ scans), and butanoic acid (in ESI-MS- scans) was also observed (Tables 1 and 2).

New On-line Mass Spectrometer for Identification of Reaction Products

Figure 6.

91

Time-profiles of some ESI-MS+ and ESI-MS- scans during the OH-oxidation of NMP in the aqueous phase.

92

L. Poulain et al.

Reaction products identified in the positive mode, and unidentified reaction products for which the MS signal was high.

Table 1.

Name

Formula

Methylamine

H3C

(M+H)+ a 32

NH2

m/z (MS-MS)+ b

O

Formamide

H2N

N-methylformamide (NMF)

46

C H O

H N

H3C

60

C H

H3 C

Acetamide

C

NH2

60

O

86

69 - 70

88

57-58

O

H

N

2-pyrrolidone (2P)

CH3

N-ethylacetamide

H2 C

H3C

H N

C

O CH3 CH3

Dimethylacetamide

H3C

N

88

C O

CH3

N

N-methylpyrrolidone (NMP)

100 58 (major) and 72 200 (NMP dimer)

O

H

Succinimide

N

O

O

100

72 (major) and 58

114

86 (major)

114

72 (major) supposed

116

88 (major) and 58

116

73

CH3

O

N-methyl succinimide (NMS)

N

O

H

O

C

1-formyl-2-pyrrolidone (FP)

N

O

CH3

5-hydroxy-N-methylpyrrolidone (5-HNMP)

HO

N

O

HO

CH2

N-hydroxymethylpyrrolidone (NHMP)

N

O

O

N-methyl-4-aminobutanoic acid

118 H3C

N H

OH CH3

N

O

O

2-hydroxy-N-methylsuccinimide

130 HO

Unidentified compound a

+

132 b

+

ESI-MS scans at (M + H); ESI-MS-MS daughter ions.

New On-line Mass Spectrometer for Identification of Reaction Products Table 2.

93

Reaction products identified in the negative mode, and unidentified reaction products for which the MS signal was high. Name

m/z (M – H + (M – H) H2O)- b

Formula

-a

H

N

O

Succinimide

97 c

O

H HO

Hydrated Succinimide

N

115 c

O

HO O

116

N-methyl-4-aminobutanoic acid H3C

N H

OH

CH3

Hydrated N-methyl succinimide (NMS)

HO

N

O

130

HO HO HO

O

C

Hydrated 1-formyl-2pyrrolidone (FP)

H

N

C

H

N

O

130

OH

or

OH

CH2(OH)

Hydrated Nhydroxymethylpyrrolidone (NHMP)

N

132

OH

OH

CH3

Hydrated 5-hydroxy-Nmethylpyrrolidone (5-HNMP)

HO

N

OH

132

OH CH3

HO

Hydrated 2-hydroxy-Nmethylsuccinimide

N

CH3

O

N

O

OH

146 OH

HO HO

or

HO

O

87

Butanoic acid OH

Unidentified compounds a

43 101 148

ESI-MS- scan at (M - H); b ESI-MS- scans of hydrated molecules at (M–H), Succinimide was identified at (M - 2H)- and its hydrate was identified at (M – 2H + H2O) after verification with standard solutions.

c

L. Poulain et al.

94 Chemical mechanisms.

Based on the results, a possible mechanism is suggested here for the OH oxidation of NMP in the aqueous phase under tropospheric conditions. The reaction probably proceeds via three different pathways (Figure 7). - Pathway (a) proceeds via hydrogen atom abstraction by the OH radical from the methyl group of NMP. This leads to the formation of an alkyl radical, which then reacts with dissolved oxygen to form a peroxy radical. By analogy with the aqueous phase behaviour of other peroxy radicals (Von Sonntag and Schuchmann, 1997), it can self-react to form a tetroxide which rapidly decomposes, leading to the formation of FP and NHMP. - Pathway (b) proceeds via hydrogen atom abstraction by OH radicals, from the CH2 group adjacent to the amine group of NMP. This leads to the formation of an alkyl radical which reacts with dissolved O2 to form another peroxy radical. This peroxy radical can also self-react to form a tetroxide, which rapidly decomposes into NMS and 5-HNMP. From the HPLC analysis a value of 55 ± 1 % can be put on the importance of this pathway. - Pathway (c) is more speculative, but it has been previously suggested by Horikoshi, et al., 2001, who investigated the OH oxidation of 2P in the aqueous phase, in the presence of TiO2. This pathway may occur for NMP under real tropospheric conditions. It proceeds via a ring-opening mechanism, leading to the formation of N-methyl-4-aminobutanoic acid. The formation of smaller molecules can be explained by the subsequent reactions of the primary products of the pathways (a), (b) or (c), which may lead to ring-opening mechanisms. CH3

N O

a

+ OH

c

b

O x

CH2

CH3

N

C

•

N

O

O

Cx

H3C

N

•

+

H2O

OH

+ O2

+ O2

C

• OO

CH3

CH2

• N

OO

O

CH2OH

CH3

N

N

FP

Figure 7.

NHMP

O

N-methyl-4aminobutanoic acid

O

N

O

O

N C

CHO

O

H3C

H N

CH3

O

NMS

HO

N O

5-HNMP

Suggested mechanism for the OH oxidation of NMP in the aqueous phase.

New On-line Mass Spectrometer for Identification of Reaction Products

95

Summary and conclusions The reactivity of NMP towards OH radicals was studied in the aqueous phase, under tropospheric conditions. The kinetic results show that the OH oxidation of NMP is fast compared to that of other WSOC, and thus should induce modifications of the composition of water droplets, due to the reaction products formed. A new experimental technique was developed to study the aqueous phase OH oxidation of NMP. A mass spectrometer was coupled to an aqueous phase simulation chamber, thus providing an on-line analysis of the solution. The mass spectrometer was equipped with an electrospray ionisation (ESI) unit and a triple quadrupole, which allowed ESI-MS+, ESI-MS-, and ESI-MS-MS+ analysis. The results proved that this experimental technique is highly promising, as it allowed us to detect the formation of 66 reaction products, of which 24 were positively identified. Based on the results obtained, a chemical mechanism has been suggested for the OH oxidation of NMP in the aqueous phase. The developed equipment can be used to study other molecules and other reactions of atmospheric interest. Acknowledgements This study was funded by the European project MOST (EVK2-CT-2001-00114). The authors also thank Séverine Barale, Marie Bonnevialle, Leïla Debiesse, Claire Detruit for their kind help with the experiments and analysis by HPLC. References Aschmann S. M. and R. Atkinson; Atmospheric chemistry of 1-methyl-2-pyrrolidinone, Atmos. Environ., 33 (1999) 591-599. Atkinson R.; Atmospheric chemistry of VOCs and NOx, Atmos. Environ., 34 (2000) 2063-2101. Campbell H. L. and B. A. Striebig; Evaluation of N-Methylpyrrolidone and its oxidation products toxicity utilizing the Microtox assay, Environ. Sci. Technol., 33 (1999) 11, 1926-1930. Claeys M., W. Wang, A. C. Ion, I. Kourtchev, A. Gelencser and W. Maenhaut; Formation of secondary organic aerosols from isoprene and its gas-phase oxidation products through reaction with hydrogen peroxide, Atmos. Environ., 38 (2004) 25, 4093-4098. Ervens B., G. Feingold, S. L. Clegg and S. M. Kreidenweis; A modeling study of aqueous production of dicarboxylic acids: 2. implications for cloud microphysics., J. Geophys. Res. [Atmos.], 109 (2004) D15, D15206/15201-D15206/15212. Ervens B., S. Gligorovski and H. Herrmann; Temperature-dependent rate constants for hydroxyl radical reactions with organic compounds in aqueous phase, Phys. Chem. Chem. Phys., 5 (2003) 1811-1824. Friesen D. A., J. V. Headley and C. H. Langford; The photooxidative degradation of N-Methylpyrrolidone in the presence of Cs3PW12O40 and TiO2 colloid photocatalysts., Environ. Sci. Technol., 33 (1999) 18, 3193 3198. Gelencser A., A. Hoffer, G. Kiss, E. Tombacz, R. Kurdi and L. Bencze; In-situ Formation of Light-Absorbing Organic Matter in Cloud Water, J. Atmos. Chem., 45 (2003) 25-33. Graedel T. E. and C. J. Weschler; Chemistry within aqueous atmospheric aerosols and raindrops, Rev. Geophys. Space Phys. 19 (1981) 505-539. Herrmann H.; Kinetics of aqueous phase reactions relevant for atmospheric chemistry, Chem. Rev. 103 (2003) 4691-4716. Hine J. and P. K. Moorkerjee; Strucural effects on rate and equilibriums. XIX. Intrinsic hydrophobic character of organic compounds. Correlations in terms of structural contributions, J. Org. Chem. 40 (1975) 293 298.

96

L. Poulain et al.

Horikoshi S., H. Hidaka and N. Serpone; Photocatalyzed degradation of polymers in aqueous semiconductor suspensions: V. Photomineralization of lactam ring-pendant polyvinylpyrrolidone at titania/water interfaces, J. Photochem. Photobiol. A: Chem. 138 (2001) 69-77. Lelieveld J. and P. J. Crutzen; Influence of cloud photochemical processes on tropospheric ozone, Nature, 343 (1990) 227-233. Monod A. and P. Carlier; Impact of clouds on the tropospheric ozone budget: Direct effect of multiphase photochemistry of soluble organic compounds, Atmos. Environ., 33 (1999) 4431-4446. Monod A., L. Poulain, S. Grubert, D. Voisin and H. Wortham; Kinetics of OH-initiated oxidation of oxigenated organic compounds in the aqueous phase: new rate constants, structure-activity relationships and atmospheric implications, Atmos. Environ. (2005) in press. Von Sonntag C. and H.-P. Schuchmann; Peroxyl radicals in aqueous solution, in: Z. Alfassi (ed.), Peroxyl radicals. John Wiley & Sons, Chichester (1997) 173-234.

Dynamic Chamber System to Measure Gaseous Compounds Emissions and Atmospheric-Biospheric Interactions Viney P. Aneja1*, Jessica Blunden1, Candis S. Claiborn2, and Hugo H. Rogers3 1

Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University , Raleigh, NC 27695-8208, U.S.A. 2 Department of Civil and Environmental Engineering, Washington State University, Pullman, Washington 99164-2910, U.S.A. 3 National Soil Dynamics Laboratory, ARS-USDA, Auburn, Alabama 36832, U.S.A.

Key Words: Dynamic chamber, Emissions, Ammonia, Nitrogen oxide, Hydrogen sulphide, Hydrogen peroxide

Abstract The dynamic flow-through chamber system has been developed in response to a need to measure emissions of nitrogen, sulphur, and carbon compounds for a variety of field applications. The cylindrical chamber system is constructed of chemically inert materials and internally lined with 5mil thick transparent fluorinated ethylene polypropylene (FEP) Teflon to reduce chemical reactions and build up of temperature inside the chamber. The chamber (diameter = 27cm, height = 42 cm, volume = 24.05 L) is designed with an open-ended bottom that can penetrate either soil or liquid to a depth of ~6-8 cm, thus creating a completely enclosed system. Carrier gas (e.g. compressed zero-grade air) is pumped at a constant flow rate (~2 to ~5 lpm), depending on the season. The air inside the chamber is well mixed by a variable-speed, motor-driven Teflon impeller (~40 to ~100 rpm). Many different laboratory and field experiments have been conducted using this dynamic chamber system. Oxides of nitrogen (NO, NO2, NOy) emissions have been measured from agricultural soils where nitrogen-rich fertilizers have been applied. Ammonia-nitrogen (NH3-N) and reduced organic sulphur compounds emissions have been measured using this same technique across a gas-liquid interface at swine waste treatment anaerobic storage lagoons, and agricultural fields. Similar chamber systems have also been deployed to measure uptake of nitrogen, sulphur, ozone, and hydrogen peroxide gases by crops and vegetation to examine atmospheric-biospheric interactions. Emissions measurements have been validated by a coupled gas-liquid transfer with chemical reaction model as well as a U.S. Environmental Protection Agency (EPA) WATER 9 model. Introduction The dynamic flow-through chamber system is a technique that has been developed and modified over the past 25 years in an effort to measure earth-atmosphere and water-atmosphere fluxes of various compounds including biogenic sulphur, oxides of nitrogen, ammonia, and methane. Determination of fluxes for these different gaseous compounds is desirable in order to permit accurate assessment of the relative roles of biogenic and anthropogenic sources in contributing to such phenomena as the atmospheric sulphate and nitrate burdens, acidity in

97 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 97–109. © 2006 Springer. Printed in the Netherlands.

98

V. P. Aneja et al

precipitation due to biogenic emissions, and contribution to aerosol formation through atmospheric photochemical processes in localized areas downwind of suspect sources. Similar systems have also been utilized to measure exposure of plants to various air pollutants and to quantify the uptake of trace gases by plants (e.g., Aneja, 1976; Rogers et al., 1979; Heck et al., 1987; Claiborn and Aneja, 1993). Hill et al. (1978) initially developed this technique to measure biogenic sulphur fluxes from a salt marsh in Long Island, New York. More recently, the dynamic flow-through chamber system has also been used to measure ammonia emissions from agricultural and non-agricultural soils (Roelle and Aneja, 2002), hog waste treatment lagoons (Aneja et al., 2000, 2001a,b), and agricultural crop soils over which liquid hog waste has been applied (Roelle and Aneja, 2002), and biogenic nitric oxide emissions from various types of agricultural soils (Li et al., 1999; Roelle et al., 2001) as well as municipal wastewater treatment plant bio-solids-amended soils (Tabachow et al., 2002). Currently, a study by the North Carolina State University Air Quality research group, in coordination with the U.S. Department of Agriculture National Research Initiative, is underway to measure simultaneous emissions of nitric oxide, ammonia and hydrogen sulphide from swine waste treatment lagoons using this same technique. For measuring emissions, the chamber technique has the important advantage of association of a particular emission site and its measurable array of physical, chemical, and microbiological properties with emissions of particular compounds or their reaction products. In addition, gas residence times in the chamber are on the order of minutes so that chemical transformations between emission and analysis may be minimized. The entire chamber system is small, lightweight and easily transportable, thus making a survey of trace gas flux very convenient. Finally, the system is relatively inexpensive and laboratory requirements are low. However, it should be noted that the chamber system cannot completely simulate the ambient environment. The chamber measures over a small area (requires extrapolation) and extrapolation to large scale may be a problem. For measuring fluxes to plants, the dynamic chamber is operated as a well-mixed chamber. Ideally, then, mixing is instantaneous, conditions are uniform throughout the chamber, and exit conditions are representative of internal, well-mixed conditions, thus simplifying calculations of fluxes and mathematical treatment of non-plant surface losses (Aneja, 1976; Claiborne and Aneja, 1993). Dynamic Flow-Through Chamber for Emissions at the Air-Water/Soil Interface In the dynamic emission flux chamber method, a continuously stirred tank reactor (CSTR; Levenspiel, 1962) with an open bottom is placed over an area of interest of soil, mud, or water (Hill et al., 1978). For flux measurements at the air-water interface, a floating platform is used to hold the chamber system above water. A hole is cut in the centre of the platform in which the chamber rests, allowing the bottom of the chamber to penetrate the water surface by 3-4 cm, thus forming a seal between the water surface and the air within the chamber and thereby providing a completely enclosed system in which to measure gaseous fluxes. Figure 1 depicts a schematic of a typical dynamic flow-through chamber system used to measure gaseous emissions from a swine waste treatment lagoon. For soil measurements, the chamber is placed onto a stainless steel ring

Dynamic Chamber System to Measure Gaseous Compounds Emissions

99

which is then inserted into the soil. Flux measurements are taken ~1 hour after the insertion of the ring into the soil to ensure steady-state conditions have been reached within the chamber. Over the years construction of the dynamic chamber has been modified. For studies conducted recently, the chamber is generally constructed as follows. The dynamic flow-through chamber system is cylindrical in nature and built from Plexiglas material. Chamber dimensions may vary but are generally about 23 cm inner diameter (i.d.) and 46 cm. in height. The entire closed system is lined on the inside with 2 mil fluorinated ethylene propylene (FEP) Teflon and stainless steel fittings in order to minimize chemical reactions with sample flow.

Meteorological Tower

Zero Air (no NHx , NOx, or Sx)

Mobile Laboratory

Continuous Duty Motor

Vent Line

Mass Flow Controller

23 cm

Outflow Impeller Stirrer

Inflow

46 cm Floating Platform

pH Probe Lagoon

Temperature Probe #1 Temperature Probe #2

Figure 1.

Schematic of dynamic flow-through chamber system configured to measure emissions from a swine waste treatment lagoon.

A sweep gas such as compressed zero-grade air is passed through a Teflon FEP sample line into the chamber. The in-flowing air is monitored by a mass flow controller and subsequently delivered into the chamber. The air inside the chamber is ideally well-mixed by a variable-speed motor-driven Teflon impeller stirrer (speeds generally range from ~40-100 rpm). The dynamic chamber system, with the continuous stirring provided by the impeller, meets the necessary criteria for performance as a continuously stirred tank reactor (CSTR). For performance as a CSTR, the chamber needs to be “ideally” mixed (Aneja, 1976). In ideal mixing, the composition of any elemental volume within the chamber is assumed to be the same as that of any other volume within the chamber. Tracer experiments (residence time distribution) have been used to test the flow and mixing characteristics of the system. The results of these mixing studies

V. P. Aneja et al

100

indicated that the dynamic chamber behaved as a “perfect” mixer with negligible stagnancy or channelling (Aneja et al., 2000). Johansson and Granat (1984) conducted research on pressure differences between the outside atmosphere and air within a chamber using a tilted water manometer, which indicated that pressure differences were below detectable limits (0.2 mm H2O). Arkinson (2003) conducted research on temperature difference between the outside atmosphere and air within the chamber. The results indicate that temperature differences are d 0.4°C. The out-flowing gas also flows through a Teflon FEP sample line and is directed into a temperature controlled mobile laboratory, which houses the analytical instruments and data acquisition system. Here, the sample is analyzed continuously for content of the gas of interest. A vent line is fitted to the exiting sample line to prevent pressurization and was periodically bubble tested to check for under pressurization and/or leaks in the enclosed system. Sample lines do not exceed 10 meters. Flux Calculations The following mass balance equation may be used for the dynamic flow-through chamber system and applied to any target gaseous substance of known concentration: dC dt where

§ LAw q · § q[Cair ] JA ·  ¸R  ¨ ¸  C ¨ V ¹ V¹ © V © V

C

concentration of gas inside the chamber (ppbV)

Cair

concentration of gas in carrier air (ppbV)

q

flow rate of compressed air through the chamber (lpm)

V

volume of the chamber (L)

A

emission surface area covered by chamber (m2)

Aw

inner surface area of the chamber of inner and upper wall

(1)

surfaces (m2) L

total loss of gas in the chamber per unit area (m min-1) due to reaction with inner and upper walls of the chamber

h

internal height of the chamber (cm)

J

emission flux per unit area (µg [gas] m-2 s-1)

R

gas phase reactions inside the chamber

Since zero-grade air is used as the carrier gas, Cair is equal to zero and gas phase reactions, R, are also assumed to be zero. Since the air inside the chamber is assumed to be well mixed by

Dynamic Chamber System to Measure Gaseous Compounds Emissions

101

the impeller stirrer, C is assumed to be constant within the chamber. At steady-state conditions, § dC · the change of concentration with respect to time ¨ ¸ is expected to be zero. Therefore equation © dt ¹ (1) can be simplified as:

J h

§ LAw q · Ceq¨  ¸ V¹ © V

(2)

The loss term, L, is determined experimentally while equilibrium-state gas concentration (Ceq), flow rate (q) and chamber dimensions (V and h) are all measured. Kaplan et al. (1988) has ª Ceq  C (t ) º devised a method for calculating loss term by calculating the slope of the plot of  ln « » ¬ Ceq  Co ¼ versus time (t). For this experiment, Co is the initial equilibrium state gaseous concentration measured by the chamber system at a constant flow rate (lpm). Ceq is the measured gaseous concentration at a second equilibrium state at an increased or reduced flow rate (lpm) into the chamber system. C(t) depicts the gaseous concentration at any time, t, during the transition between the first and second equilibrium states. L is determined by:

L

q ·§ V · § ¨ slope  ¸¨ ¸ V ¹© Aw ¹ ©

(3)

Dynamic Flow-Through Chamber for Measuring Trace Gas Uptake by Plants For measuring uptake of gaseous species by plants, and for conducting exposure experiments, it is important to characterize the conditions very near the plant surface under study; thus, assuring well-mixed conditions for such experiments becomes very important. Claiborn and Aneja (1993) constructed a chamber with a 1:1 diameter to height ratio, which is ideal for wellmixed conditions (Uhl and Gray, 1966). The chamber is constructed of a Teflon-coated strap iron frame lined internally with a 5 mil thick Teflon film to allow for maximum light, and even light and temperature distribution. Mixing is provided by a Teflon-coated impeller mounted at the top of the frame, and aided by 3 Teflon-coated baffles. The entire system operates under slight negative pressure, and it is important to assure that there are no leaks into the system, to allow for proper determination of the air flow rate through the system and the initial concentration of the trace gas of interest. The air flow rate through the chamber provides one to two chamber exchange rates per minute. Sample ports are located on the inlet and outlet Teflon pipes. Separate sample manifolds are provided for CO2/water vapour and the trace gas of interest. Air samples from the chamber inlet and outlet are alternated automatically to the analyzers so that concentration differences across the chamber are determined using the same analyzer. In order to determine trace gas uptake, wall losses to the empty chamber surfaces must first be determined. The fraction of the trace gas introduced into the chamber that is lost to the chamber walls during the chamber loss experiment (CL) is determined:

V. P. Aneja et al

102

CL

C Ao ,empty  C A,empty C Ao ,empty

(4)

where CA and CAo are the outlet and inlet concentrations, respectively, of trace gas species A, and the subscript empty indicates the concentrations measured for the empty chamber experiment. Total losses (i.e., to both plant surfaces and chamber surfaces) are then measured, and the fractional loss (TL) of the trace gas entering the chamber is similarly determined:

TL

C Ao ,exp  C A,exp

(5)

C Ao ,exp

where CA and CAo are, again, the outlet and inlet concentrations, respectively, of trace gas species A, and the subscript exp indicates the concentrations measured for the exposure chamber experiment. Treatment of Wall Losses For trace gases that are not very reactive, it is sufficient to subtract the wall losses from the empty chamber experiment from the total trace gas losses from the plant uptake experiments in order to determine the losses to plant surfaces, alone. In other words, BL = TL-CL

(6)

where BL is the fractional loss of the trace gas to plant biomass surfaces, alone. For very reactive and/or water-soluble species like hydrogen peroxide or nitric acid, however, a more rigorous method of correcting for wall losses must be employed (Claiborn and Aneja, 1993). By treating both wall losses and losses to plant surfaces as two parallel reactions, the fractional loss to the plant biomass surfaces, BL, is calculated from TL and CL,

BL

TL  CL 1  CL

(7)

From this equation, it is evident that for species for which the chamber losses are very low, it is acceptable to simply subtract the fractional loss to the walls (CL) from the fractional loss during the exposure chamber (TL). Finally, fluxes to plant surfaces (QA) are calculated from the fractional loss to the plant surfaces (BL), the chamber air flow rate (F), the inlet trace gas concentration, and the plant surface area (AL): QA

BL F C Ao,exp AL

(8)

Dynamic Chamber System to Measure Gaseous Compounds Emissions

103

Results The graphs in Figure 2 (a,b,c,d) provide results from some field studies undertaken to measure fluxes of selected S, N, and C compounds. Figure 2a (Aneja et al., 2001c) reports the log of hourly averaged ammonia-nitrogen flux from a swine waste treatment lagoon plotted against lagoon aqueous phase surface temperature during four different seasonal measurements. The graph shows an observed exponential (r2 = 0.78) relationship between NH3 and lagoon water temperature measured over the year. Figure 2b demonstrates the effect of soil temperature on NO from an agricultural (soybean) crop soil at four different sites. Figure 2c illustrates the temperature dependence of average H2S emission fluxes from a salt marsh. The emission rate appears to increase semi-logarithmically with increasing temperature within a temperature range of ~0qC to ~40°C. Finally, Figure 2d represents isoprene (hydrocarbon) concentration plotted against temperature during warm months (April through September). Please note that concentrations (ppbV), which are treated as a surrogate for emission (since the concentration measurements are in close proximity to the source), are reported rather than flux calculations. The graph depicts the linear relationship between the logarithm of isoprene concentration and temperature using data from four sites.

a)

(c)

Figure 2 (a,b,c,d).

(b)

(d)

Emissions of NH3 (Aneja et al., 2000), NO (Aneja et al., 2001c), Isoprene (Hagerman et al., 1997), and H2S (Hill et al., 1978), respectively, versus Temperature at various locations.

V. P. Aneja et al

104

The well-mixed chamber was used to measure uptake of gaseous hydrogen peroxide by spruce saplings (coniferous species). Hydrogen peroxide is a highly reactive and water-soluble atmospheric oxidant, for which wall losses are very significant (10-35% for 1-2 air exchanges per minute) and highly variable (standard deviation of 50% for daytime experiments) (Claiborn, 1991). A linear relationship between hydrogen peroxide flux and chamber concentration was apparent for the red spruce saplings (Figure 3). Note that there is no significant difference between daytime and night time fluxes. Average deposition velocities were calculated for these experiments, and ranged from 0.08 to 0.12 cm/s, and 0.08 – 0.12 cm/s for daytime and night time values, respectively. In contrast, similar uptake experiments conducted on bean plants (broad leaf plants) resulted in deposition velocities ranging from 0.8 to 1.4 and 0.5 to 0.6 for days and nights, respectively (Claiborn, 1991). H2O2 uptake by Red Spruce saplings 25

Day Night

Flux, x 10^15 mol/cm2-s

20

15

10

5

0 0

1

2

3

4

5

6

7

Chamber concentration, ppbv

Figure 3 .

Fluxes of gaseous H2O2 to needles of red spruce saplings as measured in wellmixed chamber experiments. Diamonds denote daytime fluxes and squares denote night time fluxes (Claiborn, 1991).

Data Comparisons As with any type of experimental procedure, it is desirable to compare data acquired in field studies with controlled laboratory studies and/or related models. Various comparisons with models, as well as other measurement techniques, have been made with data collected via the dynamic flow-through chamber system. Some results are presented in the following discussion. For model comparisons with the dynamic flow-through chamber system, Aneja et al., (2001d) developed a mass transport model based on the quiescent thin film concept (Danckwerts,

Dynamic Chamber System to Measure Gaseous Compounds Emissions

105

1970), which takes into account molecular diffusion and chemical reactions. Figure 4 shows seasonal variation comparison of ammonia emissions between modelling results and dynamic chamber system experiments for ammonia emissions from a swine waste treatment lagoon from a field study conducted by Aneja et al. (2000). The coupled mass transfer model simulations corroborated experimental results utilizing the dynamic chamber technique and validated the application to the flux equation. Also, Figure 5 shows how closely the recently developed WATER9 model, based on a set of air emission models that are documented in Air Emissions Models for Waste and Wastewater, U.S. Environmental Protection Agency, (U.S. EPA, 1994), compare with the same set of ammonia emissions data as discussed with the mass transport model previously. Tabachow et al. (2002) conducted and compared the results of nitric oxide (NO) emissions measurements for both in situ field and laboratory experiments from unamended and municipal wastewater treatment plant biosolids-amended soil. Based on matching soil temperatures (within 5ºC) and water filled pore space (WFPS) (± 5%), the ratio of volumetric soil water to total porosity of the soil, five of six scenarios showed no statistically significant difference in the NO flux measurements from the laboratory versus the field studies.

Figure 4 .

Seasonal variation comparison of ammonia emissions between modelling results and dynamic chamber system experiments.

Nitric oxide emissions from soils on a soybean field, measured by the dynamic chamber system (Li et al., 1999), were compared in Figure 6 with NO fluxes calculated at a 5 meter height utilizing the eddy-correlation method (Gao et al., 1996). In spite of differences in flux magnitudes (note the different scales used), the two show qualitatively similar variations and structure with time. The values of NO soil emissions are greater than the NO fluxes at 5 m, which implies that

V. P. Aneja et al

106

some of the NO might have been converted to nitrogen dioxide (NO2) through reactions with ozone and peroxy radicals by the time it reaches the 5-m height level from the source. In a coordinated research effort (Project NOVA), NO emissions were measured from fine sandy loam soil in a corn field using two different (dynamic and static) chamber techniques. One group from North Carolina State University used the dynamic chamber method while a group from NASA Langley Research Center used a closed box flux technique (static chamber), in which NO fluxes were calculated using the mixing ratio of NO (ppbv) versus time. (Anderson and Levine, 1987). Figure 7 depicts a difference plot of the composite averaged fluxes calculated by the dynamic chamber technique and the static chamber technique. Upon, statistical analysis, it was shown that there was no statistically significant difference between NO flux as measured using the two different chamber methods. Chamber experiments on uptake by plants provide useful information for the determination of deposition velocities, however, the deposition velocities over forest canopies will be different to those calculated in the chamber experiments. Deposition velocities take into account not only the resistance to mass transfer due to the plant itself, but also the resistance due to the forest canopy structure, and the turbulence over the canopy.

U.S. EPA WATER9 Model Ammonia Flux Measurements

Figure 5 .

Ammonia flux versus lagoon temperature comparison between WATER9 modelling results and dynamic chamber experimental calculated results.

Dynamic Chamber System to Measure Gaseous Compounds Emissions

107

Figure 6 .

Inter-comparison of NO fluxes at 5 meter height using eddy-correlation technique with NO soil emissions using dynamic chamber technique.

Figure 7.

Difference of NO flux between two chamber techniques (dynamic chamber flux – static chamber flux) versus time of day. Vertical lines indicate one standard deviation of the NO flux measurements made from both chamber techniques.

Conclusions The dynamic flow-through chamber system has been successfully developed in response to a need to measure emissions of nitrogen, sulphur, and carbon compounds for a variety of field applications. Moreover, similar chamber systems have also been deployed to measure uptake of nitrogen, sulphur, ozone, and hydrogen peroxide gases by crops and vegetation to examine

108

V. P. Aneja et al

atmospheric-biospheric interactions. Experimental results have been effectively validated by independent flux models. Acknowledgements Funding of this research project was provided by the U.S. Department of Agriculture (USDA) as a part of the National Research Initiative (NRI) under Contract No. 2003-05360. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. References Anderson, I.C. and J.S. Levine, Simultaneous field measurements of biogenic emissions of nitric oxide and nitrous oxide, J. Geophys. Res., 92, 965-976, 1987. Aneja, V.P., Dynamic Studies of Ammonia Uptake by selected Plant Species Under Flow Reactor conditions, Ph.D. Thesis, North Carolina State University, Raleigh, N.C., pp. 216, 1976. Aneja, V.P., G.C. Murray, and J. Southerland, Atmospheric nitrogen compounds: Emissions, transport, deposition, and assessment, EM, April 22-25, 1998. Aneja, V.P., J.P. Chauhan, and J.T. Walker, Characterization of atmospheric ammonia emissions from swine waste storage and treatment lagoons, J. Geophys. Res., 105, No. D9, 11535-11545, 2000. Aneja, V.P., B. Bunton, J.T. Walker, and B.P. Malik, Measurement and analysis of atmospheric ammonia emissions from anaerobic lagoons, Atmos. Environ., 35, 1949-1958, 2001a. Aneja, V.P., P.A. Roelle, G.C. Murray, J. Southerland, J.W. Erisman, D. Fowler, W.A.H. Asman, and N. Patni, Atmopheric nitrogen compounds II: emissions, transport, transformation, deposition, and assessment. Atmos. Environ., 35, 1903-1911, 2001b. Aneja, V.P., P.A. Roelle, and Y. Li, Effect of environmental variables on NO emissions from agricultural soils, Phyton (Austria), 4, 29-40, 2001c. Aneja, V.P., B.P. Malik, Q. Tong, D. Kang, and J.H. Overton, Measurement and Modelling of ammonia emissions at waste treatment lagoon-atmospheric interface, Water, Air, and Soil Pollution, 1, 177-188, 2001d. Arkinson, H.L., Measurements, Modeling, and Analysis of Ammonia Flux from Hog Waste treatment technologies, M.S. Thesis, pp.129, 2003. Claiborn, C.S., Atmospheric-Biospheric Interactions of Ambient H2O2: Climatology at Mt. Mitchell, N.C., and Transport into Needles of Red Spruce, Ph.D. Thesis, North Carolina State University, 1991. Claiborn, C. S., and V.P. Aneja. Transport and fate of reactive trace gases in red Spruce needles, part 1: Uptake of gaseous hydrogen peroxide as measured in controlled chamber flux experiments. Environmental Science and Technology 27, 2585-2592, 1993. Danckwerts, P.V., Gas-Liquid Reactions, McGraw-Hill Book Company, New York, 1970. Gao, W., M.L. Wesely, D.R. Cook, and T.J. Martian, Eddy-correlation measurements of NO, NO2 and O3 fluxes, in Proceedings of Measurement of Toxic and Related Air Pollutants, pp. 146-150, Air and Waste Manage. Assoc., Pittsburgh, PA, 1996. Hagerman, L.M., V.P. Aneja, and W.A. Lonneman, Characterization of non-methan hydrocarbons in the rural southeast United States, Atmos. Environ., 31, 4017-4038, 1997.

Dynamic Chamber System to Measure Gaseous Compounds Emissions

109

Heck, W.W.,O.C. Taylor, and D.T. Tingey, Assessment of Crop Loss from Air Pollutants, Elsevier Applied Science, London, U.K. p. 552, 1987. Hill, F.B, V.P. Aneja, and R.M. Felder, Atechnique for measurement of biogenic sulfur emission fluxes, J. Environ. Sci. Health, A13(3), 199-225, 1978. Kim, D.S., V.P. Aneja, and W.P. Robarge, Characterization of nitrogen oxide fluxes from soil of a fallow field in the coastal piedmont of North Carolina, Atmos. Environ., 28, 1129-1137, 1994. Levenspiel, O., Chemical Reaction Engineering, John Wiley &Sons, New York, 1962. Li, Y., V.P. Aneja, S.P. Arya, J. Rickman, J. Brittig, P. Roelle, and D.S. Kim, Nitric oxide emissions from intensively managed agricultural soil in North Carolina, J. Geophys. Res., 104, 26,115-26,123, 1999. Roelle, P.A. and V.P. Aneja, Characterization of ammonia emissions from soils in the upper coastal plain, North Carolina, Atmos. Environ., 36, 1087-1097, 2002. Roelle, P.A., V.P. Aneja, B. Gay, C. Geron, and T. Pierce, Biogenic nitric oxide emissions from cropland soils, Atmos. Environ., 35, 115-124, 2001. Roelle, P.A., V.P. Aneja, J. O’Connor, W. Robarge, D.S. Kim, and J.S. Levine, Measurement of nitrogen oxide emissions from an agricultural soil with a dynamic chamber system, J. Geophys. Res., 104, 1609-1619, 1999. Rogers, H.H., H.E. Jeffries, and A.M. Witherspoon. Measuring air pollutant uptake by plants: nitrogen dioxide. Journal of Environmental Quality 8, 551-557, 1979. Tabachow, R.M., P.A. Roelle, J.J. Pierce, and V.P. Aneja, Soil nitric oxide emissions: lab and field measurements and comparison, Environmental Engineering Science, 19(4), 205-213, 2002. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, November 1994, EPA-453/R-94-080A.

Chamber Studies on the Photolysis of Aldehydes John C. Wenger Department of Chemistry and Environmental Research Institute, University College Cork, Ireland Key Words: Photolysis, Aldehydes, Photooxidation

Introduction Aldehydes are emitted directly into the atmosphere from a variety of natural and anthropogenic sources and are also formed in situ from the atmospheric degradation of volatile organic compounds (VOCs). The atmospheric fate of aldehydes is controlled by photolysis and reaction with hydroxyl (OH) or nitrate (NO3) radicals and, in the case of unsaturated compounds, reaction with ozone (Atkinson, 1994). The photolysis of aldehydes is of particular importance because it is a source of free radicals in the troposphere, and thus may significantly influence the oxidizing capacity of the lower atmosphere (Finlayson-Pitts and Pitts, 1986). Aldehydes absorb in the near ultraviolet range and typically exhibit a weak absorption band in the wavelength range 240-360 nm as a result of a symmetry forbidden n-S* transition. The minimum wavelength of sunlight that reaches the troposphere is 290 nm, which is of sufficient energy to cause photolysis of aldehydes. The importance of photolysis relative to other loss processes such as reaction with OH, NO3 or O3 can be determined by comparison of the rates of all the competing degradation pathways. If photolysis is found to be important then the products need to be identified and mechanisms constructed in order to provide a more complete picture of the photolysis process. In this paper an overview of the application of simulation chamber studies to the photolysis of aldehydes is presented and a brief review of the kinetic and mechanistic data currently available is also provided. Finally, results from some of the first studies of the photolysis of aromatic aldehydes in an outdoor simulation chamber are presented and discussed. Determination of Photolysis Rate Coefficients The photolysis of a species X follows first order kinetics; j X + hQ o Products

and is described by the photolysis rate coefficient j. The rate of removal of X by photolysis is thus; -d[X]/dt = j[X] The photolysis rate coefficient (in s-1) depends on the absorption cross-section (ı, base e in cm2 molecule-1), the quantum yield for photolysis (ij, in molecule photons-1) and the solar actinic flux (F, in photons cm-2 s-1), all of which are wavelength dependent; j = ı(Ȝ) ij(Ȝ) F(Ȝ) 111 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 111–119. © 2006 Springer. Printed in the Netherlands.

(,)

J. C. Wenger

112

Under atmospheric conditions where a range of wavelengths at Ȝ > 290 nm are present then the atmospheric photolysis rate coefficient is best described by the following expression; Ȝi

j

¦ ı(Ȝ)M(Ȝ) F (O )

(II)

Ȝ 290 nm

where ı(Ȝ) and ij(Ȝ) are averaged over the wavelength interval ǻȜ and F is summed over the same wavelength range. The sum is carried out from 290 nm to some wavelength Ȝi at which either the quantum yield or absorption cross-section becomes negligible (Finlayson-Pitts and Pitts, 1986). Calculation of j requires the values of ı and ij to be known. While the experimental determination of absorption cross-sections is fairly straightforward, the measurement of quantum yields is not, mainly due to the rapid secondary reactions of radicals. In fact, there is no quantum yield data available for the vast majority of aldehydes. As a result, calculations of photolysis rates are often carried out in which it is assumed that ij(Ȝ) = 1.0. This, of course, only produces a value for the maximum photolysis rate. In addition, because the actinic flux depends on many factors, including geographical location, time and season, photolysis rates are normally calculated for certain conditions; e.g. at noon on a cloudless day at a location 40°N on the Earth’s surface on July 1. The problems associated with calculating photolysis rates can be overcome by the experimental determination of j values in outdoor simulation chambers, such as the European Photoreactor (EUPHORE), in Valencia, Spain (Becker, 1996). The decay of an aldehyde when irradiated by natural sunlight can be measured directly by FTIR spectroscopy or gas chromatography and the j values determined from a simple first order kinetic plot in the form of equation (III); ln[X]0/[X]t = -jt

(,II)

To ensure that photolysis is the only loss process for the aldehyde experiments can be carried out in the presence of an excess concentration of a radical scavenger such as cyclohexane. In cases where the high concentration of a scavenger is undesirable, e.g. because it causes saturation in the infrared absorption spectrum, a tracer compound, such as di-n butyl ether, can be used to correct for the measured decay of the aldehyde in order to obtain the j value. Typical starting concentrations used in photolysis experiments at EUPHORE are [aldehyde] = 0.5-1.5 ppmv, [scavenger] = 10-50 ppmv or [tracer] = 0.1-02. ppmv (Wenger et al., 2004 and Magneron et al., 2002). The experimentally determined value of j is, of course, only measured under one particular set of atmospheric conditions. Since the photolysis rate depends on the intensity and spectral distribution of the sunlight it is useful to quote the photolysis rate coefficient relative to a conventional measure of solar light intensity, such as j(NO2). The value for j(aldehyde)/j(NO2) can be used to calculate the photolysis rate of the aldehyde under a range of atmospheric conditions. Information on the quantum yield for photolysis of the aldehyde by sunlight can de obtained by comparison of the experimentally determined value of j with the maximum theoretical value. The latter is calculated from solar flux intensity measurements during the experiment, known absorption cross-section data and assuming ij(Ȝ) = 1.0. The ratio of the

Chamber Studies on the Photolysis of Aldehydes

113

measured j value to the calculated value gives the effective quantum yield, ijeff, over the atmospheric absorption range of the aldehyde. During the last 5 years, the photolysis of over twenty aldehydes has been performed at EUPHORE. Many of these studies were performed as part of the EU-funded research project RADICAL (Moortgat, 2000). A summary of the reported j values and effective quantum yields is provided in Table 1 along with the calculated lifetimes due to photolysis (IJphot = 1/jexp) and reaction with OH radicals (IJOH = 1/kOH[OH]). Table 1.

Summary of photolysis rate coefficients (jexp) measured at EUPHORE. IJphot (hr)

IJOHa (hr)

(2.9 ± 2.7) × 0.06 ± 0.05 10-6

96

17

(1.1 ± 0.1) × 10-5

3.8 × 10-5

0.28 ± 0.04

29

14

Butyraldehyde

(1.0 ± 0.2) × 10-5

5.1 × 10-5

0.20 ± 0.04

29

12

i-Butyraldehyde

(3.7 ± 0.1) × 10-5

5.2 × 10-5

0.71 ± 0.02

7.5

11

n-Pentanal

(1.6 ± 0.2) × 10-5

5.4 × 10-5

0.30 ± 0.02

17

10

-5

-5

0.72 ± 0.03

7.3

12

3-Methylbutyraldehyde

(1.25 ± 0.1) × 10

-5

0.27 ± 0.01

22

12

Pivaldehyde

(1.45 ± 0.1) × 10-5 2.6 × 10-5

0.56 ± 0.05

19

10

n-Hexanal

(1.65 ± 0.3) × 10-5 5.9 × 10-5

0.28 ± 0.05

17

15

n-Nonanal

(1.15 ± 0.2) × 10-5 4.9 × 10-5

0.23 ± 0.03

24

10

Glyoxal

(1.05 ± 0.3) × 10-4 2.7 × 10-3

0.04 ± 0.01

2.6

25

Glycolaldehyde

(1.15 ± 0.3) × 10-5 8.6 × 10-6

Compound

jexp (s-1)

jmax (ij = 1) (s-1)

Acetaldehyde

(2.9 ± 2.7) × 10-6

Propionaldehyde

2-Methylbutyraldehyde

Methacrolein

(3.8 ± 0.1) × 10

-5

-6

< × 10

-6

5.2 × 10

4.7 × 10

ijeff

1.32 ± 0.30

24

28

5.2 × 10

-4

< 0.004

> 144

8

-4

< 0.004

> 144

13 7.4

Acrolein

< × 10

4.3 × 10

trans-Crotonaldehyde

(1.2 ± 0.2) × 10-5

4.0 × 10-4

0.03 ± 0.01

24

Pinonaldehyde

(1.15 ± 0.1) × 10-5 8.0 × 10-5

0.14 ± 0.01

24

3.1 6

a) OH rate coefficients taken from (Atkinson, 1994), lifetime calculated assuming [OH] = 1 × 10 molecule cm-3.

The straight chain C3-C9 aldehydes all have similar photolysis rates and values for ijeff in the range 0.20-0.30. For these compounds the atmospheric lifetimes for photolysis by sunlight are longer than the lifetimes for reaction with OH radicals. In contrast, the Įbranched aldehydes possess significantly higher values for ijeff and their photolysis lifetimes are shorter than those for reaction with OH radicals. Interestingly, the ijeff values for the

J. C. Wenger

114

unsaturated aldehydes, methacrolein, acrolein and trans-crotonaldehyde are very small, even though these compounds possess absorption spectra extending towards the visible. In many atmospheric chemistry models quantum yields are assumed to be unity, which, as shown in Table 1, is seldom the case. Thus it appears that the importance of photolysis in the degradation of the aldehydes and their contribution to radical formation in the atmosphere is often over-estimated. However, details concerning the production of radicals from the atmospheric photolysis of aldehydes are best obtained from mechanistic studies. Mechanisms for the Photolysis of Aldehydes There are a number of theoretical reaction pathways in the photolysis of aldehydes, as demonstrated below for n-hexanal (Tadiü et al., 2001); CH3CH2CH2CH2CH2CHO + hȞ ĺ ĺ ĺ ĺ ĺ ĺ

n-C5H11 + HCO (Norrish Type I) n-C5H12 + CO C4H8 + CH2=CHOH (Norrish Type II) C3H7 + CH2CH2CHO C5H10 + HCHO CH3CH2CH2CH2CH2CO + H

(1) (2) (3) (4) (5) (6)

Reactions (1) and (3) are believed to be the most important under atmospheric conditions. The Norrish Type I process results in fragmentation into free radicals, whilst the Norrish Type II process, which is common to molecules with a Ȗ-hydrogen atom, is an intramolecular rearrangement that results in no radical formation. Experiments performed at EUPHORE and in indoor photoreactors as part of the RADICAL project found that the Norris Type I process is dominant for n-butanal and smaller straight chain aldehydes, but the Norrish Type II process is the major pathway for the photolysis of n-pentanal and n-hexanal. Simulation Chamber Studies on the Photolysis of Aromatic Aldehydes Aromatic aldehydes are emitted into the atmosphere as primary pollutants from automobile exhausts and are also formed in situ from the OH radical initiated oxidation of alkyl aromatic compounds. Tolualdehydes and dimethylbenzaldehydes are produced from the gas-phase reaction of OH with xylenes and trimethylbenzenes respectively (Calvert et al., 2002). The subsequent atmospheric fate of these aromatic aldehydes is controlled by reaction with OH and NO3 radicals and photolysis by sunlight. Kinetic studies of the gas-phase reactions of the tolualdehydes and dimethlybenzaldehydes have been recently performed (Thiault et al., 2002 and Clifford, 2004) and preliminary studies of the photolysis of the tolualdehydes have also been reported (Volkamer et al., 2000 and Thiault et al., 2001). In this work the photolysis of the 3 tolualdehydes and 6 dimethylbenzaldehydes (DMBAs) were investigated during July 2003. The measured photolysis rate coefficients have been used to calculate tropospheric lifetimes and provide an assessment of the relative importance of photolysis as an atmospheric loss process for the compounds. Experiments were performed in Chamber B of the EUPHORE facility. Reactants were introduced into the chamber by gently heating a known amount of the compound in a glass impinger and flushing the vapour into the chamber via a stream of purified air. Concentrations ranging from 131 to 511 ppbv of reactant were used during different experiments. The

Chamber Studies on the Photolysis of Aldehydes

115

reactant was allowed to mix for approximately one hour and its rate of deposition to the walls of the chamber was measured during this period. Photolysis was then initiated by opening the protective housing and thereby exposing the chamber to sunlight. The solar actinic flux was measured over the range 290-520 nm with a spectral resolution of 1 nm FWHM using a calibrated spectroradiometer (Bentham TM300). A full spectral scan took 420-430 seconds. The duration of photolysis was typically 2-3 hours. The temperature inside the chamber increased slightly as the experiments progressed but was always within the range 296-308 K. Chemical analysis was performed throughout the experiments by in situ FTIR spectroscopy using an optical path length of 553.5 m. Infrared spectra in the range 600-4000 cm-1 were obtained using a Nicolet Magna 550 FTIR spectrometer operated at a resolution of 1 cm-1 using a broad band MCT detector and derived from the co-addition of 270 scans collected over 5 minutes. Additional chemical detection was provided by a gas chromatograph (Fisons 8160) equipped with flame ionisation and photoionisation detectors (FID and PID). The chromatograph was operated using a 30 m DB-624 fused silica capillary column (J&W Scientific, 0.32 mm i.d., 1.8 Pm film). Air was sampled from the chamber into a 3 ml sampling loop and then injected onto the column. The temperature of the column was held isothermal (120º C for tolualdehydes and 150ºC for dimethylbenzaldehydes). The reactants were quantified using calibrated reference infrared spectra and gas chromatographic sensitivity factors obtained by introducing known volumes of pure materials into the chamber. The leak rate from the chamber was determined daily by adding about 20 ppbv of the unreactive tracer gas SF6 and measuring its loss by FTIR spectroscopy. The derived correction factors were applied to determine the amounts of reactants consumed and products formed. Table 2.

Experimental details for the photolysis of 2,4-DMBA at EUPHORE. Date

9th July 2003

Initial Concentration (ppbv) Irradiation time j (NO2)average (s-1) Results k(2,4-DMBA)FTIR (s-1) k(2,6-DMBA)GC-PID (s-1) kSF6 (s-1) kwall (s-1) j(2,4-DMBA)FTIR (s-1) j(2,4-DMBA)GC-PID (s-1) j(2,4-DMBA)average (s-1) j(2,4-DMBA)/j(NO2))

303 2hr 39min (7.47 ± 0.75) × 10-3 (2.14 ± 0.02) × 10-4 (2.43 ± 0.04) × 10-4 (9.19 ± 0.88) × 10-6 (2.44 ± 0.76) × 10-5 (1.90 ± 0.03) × 10-4 (2.19 ± 0.03) × 10-4 (2.05 ± 0.03) × 10-4 (2.74 ± 0.27) × 10-2

Except for j(NO2), quoted errors are twice the standard deviation arising from the least squares fit of the data. For j(NO2) the estimated error is 10%.

A total of eight experiments were performed at EUPHORE in July 2003. The experiments were conducted for at least two hours during the middle of the day. The loss of the aromatic aldehydes was monitored using FTIR spectroscopy and GC-PID. The measured decay rates of the compounds were corrected for wall loss and dilution. Concentration-time

J. C. Wenger

116

profiles and kinetic plots were generated for each of experiments. Data for the photolysis of one of the compounds, 2,4-dimethylbenzaldehyde (2,4-DMBA), are shown in Figures 1 and 2, with the associated experimental details listed in Table 2. Similar data were obtained for all the aromatic aldehydes. A summary of the photolysis rate coefficients obtained in the EUPHORE experiments is given in Table 3. The values quoted for j(aldehyde) are the average of the data obtained by FTIR spectroscopy and GC-PID in several experiments. The ratio j(aldehyde)/j(NO2) was also determined and can be used to calculate the photolysis rate of the aromatic aldehydes under different light conditions in the real atmosphere. The effective quantum yields obtained for o-, m- and p-tolualdehyde were determined by comparing the measured j values with the maximum theoretical values. The latter were calculated from solar flux intensity measurements during the experiments, known absorption cross-section data (Thiault et al., 2004) and assuming ij(Ȝ) = 1.0. There is no gas-phase absorption cross-section data available for the dimethylbenzaldehydes. However, inspection of the solution phase spectra indicates that the light absorption characteristics will be similar to those of the tolualdehydes. From the measured experimental photolysis rates of 2,3-, 2,4-, 2,5- and 2,6- DMBA it is expected that the effective quantum yields will be similar to that of o-tolualdehyde and will be close to unity. For 3,4- and 3,5DMBA, which did not photolyse at EUPHORE, it is expected that the quantum yields will be similar to m- and p-tolualdehyde, i.e. about 0.03.

350

0.008

Chamber Closed

Chamber Open

0.007

250 0.006 200

0.005

150

0.004

100

FTIR GC-PID j(NO2)

0.003 0.002

50

0.001

0 08:24

j(NO2) (s-1)

2,4-DMBA (ppbv)

300

0.009

0.000 08:52

09:21

09:50

10:19

10:48

11:16

11:45

12:14

Time of day (hh:mm)

Figure 1.

Concentration-time profile and j(NO2) during the photolysis of 2,4-DMBA at EUPHORE on 9th July 2003.

The following compounds were readily photolysed by sunlight; o-tolualdehyde, 2,3-, 2,4-, 2,5- and 2,6-DMBA. The photolysis rate coefficients are all remarkably similar, with values for j(aldehyde) in the range 2.0 – 2.7 × 10-4 s-1. For the compounds that did not undergo photolysis (m- and p-tolualdehyde, 3,4- and 3,5- DMBA), the rate of decay of the compounds is equivalent to the wall loss or dilution rate. The j values determined for these

Chamber Studies on the Photolysis of Aldehydes

117

compounds are therefore upper limits. The photolysis rate coefficients obtained for o-, m- and p-tolualdehyde in this study are in good agreement with those previously reported (Volkamer et al., 2000 and Thiault et al., 2001). As this is the first study on the photolysis of the DMBAs, no comparative data exists. 0

ln[2,4-DMBA]0/[2,4-DMBA]t - kwallt

-0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4 -1.6

FTIR GC-PID

-1.8 -2 0

1000

2000

3000

4000

5000

6000

7000

8000

9000

Photolysis time (s)

Figure 2.

Photolytic loss rate of 2,4-DMBA at EUPHORE on 9th July 2003.

Table 3.

Results obtained for the photolysis of the tolulaldehydes and DMBAs at EUPHORE.

Compound

o-tolualdehyde m-tolualdehyde p-tolualdehyde 2,3-DMBA 2,4-DMBA 2,5-DMBA 2,6-DMBA 3,4-DMBA 3,5-DMBA

j(aldehyde) (s-1)

j(aldehyde)/j(NO2) -4

(2.15 ± 0.06) × 10 (7.43 ± 0.60) × 10-6 (7.43 ± 0.60) × 10-6 (2.28 ± 0.20) × 10-4 (2.05 ± 0.03) × 10-4 (2.36 ± 0.09) × 10-4 (2.31 ± 0.06) × 10-4 (9.36 ± 0.20) × 10-6 (8.27 ± 0.18) × 10-6

-2

(2.66 ± 0.27) × 10 (0.09 ± 0.01) × 10-2 (0.09 ± 0.01) × 10-2 (3.02 ± 0.30) × 10-2 (2.74 ± 0.27) × 10-2 (3.52 ± 0.35) × 10-2 (3.01 ± 0.30) × 10-2 (0.11 ± 0.01) × 10-2 (0.10 ± 0.01) × 10-2

ijeff

1.19 ± 0.04 < 0.05 < 0.03 1.0* 1.0* 1.0* 1.0* 0.03* 0.03*

* Estimated as no gas-phase data available The photolysis rate coefficients obtained in this work can be used to calculate the tropospheric lifetimes of the aromatic aldehydes with respect to photolysis. The lifetimes listed in Table 4 were calculated from the values of j(aldehyde)/j(NO2) obtained at

J. C. Wenger

118

EUPHORE and assuming a value of j(NO2) = 1 × 10-2 s-1, which is typical of the maximum photolysis rate at solar noon during summer months. The atmospheric lifetimes due to reaction with OH and NO3 are also shown for comparative purposes. For o-tolualdehyde, 2,3-, 2,4-, 2,5- and 2,6-DMBA, photolysis is clearly the dominant atmospheric loss process, whilst for the remaining aldehydes, gas-phase reaction with OH radicals is the most important degradation pathway. Lifetimes (in hours) of the aromatic aldehydes with respect to the various atmospheric loss processes.

Table 4.

Compound

WOH

WNO3

Wphotolysis

o-tolualdehyde m-tolualdehyde p-tolualdehyde 2,3-DMBA 2,4-DMBA 2,5-DMBA 2,6-DMBA 3,4-DMBA 3,5-DMBA

8.7 8.4 8.5 6.8 6.4 6.3 5.8 7.1 6.2

56.7 58.7 58.7

1.0 30.9 30.9 0.9 1.0 0.8 0.9 25.3 27.8

The OH and NO3 lifetimes were calculated from rate coefficients determined by (Clifford, 2004). No rate coefficients are available for the reaction of NO3 with the DMBAs.

The fact that photolysis only occurs for compounds in which the -CHO group is adjacent (ortho) to a methyl group suggests that some form of intramolecular interaction between these functional groups is a significant component of the photolysis mechanism. This aspect is currently being investigated further and photolysis mechanisms are in development. Acknowledgements This author would like to thank the following co-workers; G.M. Clifford, A. Mellouki, G. Le Bras, A. Munoz, M. Martín-Reviejo and K. Wirtz. This work was supported by the European Commission, through the IALSI project (Contract EVR1-CT-2001-40013), the Irish Higher Education Authority (HEA) and the French Programme National de Chimie Atmosphérique (PNCA). The CEAM Foundation is supported by the Generalitat Valenciana and BANCAIXA. References Atkinson, R; Gas-phase tropospheric chemistry of organic compounds, J. Phys. Chem. Ref. Data, Monograph 2 (1994) 1-216. Becker, K.H. (Ed.); The European Photoreactor EUPHORE, Final Report of the EC Project EV5V-CT92-0059, Wuppertal (1996). Clifford, G.M.; Atmospheric chemistry of aromatic aldehydes, PhD thesis, University College Cork (2004). Finlayson-Pitts, B.J., J.N. Pitts Jr.; Atmospheric Chemistry, John Wiley, New York (1986) Magneron, I., R. Thévenet, A. Mellouki, G. Le Bras, G.K. Moortgat and K. Wirtz; A study of the photolysis and OH-initiated oxidation of acrolein and trans-crotonaldehyde, J. Phys. Chem. A 106 (2002) 2526.

Chamber Studies on the Photolysis of Aldehydes

119

Moortgat, G.K. (Ed.); Evaluation of radical sources in atmospheric chemistry through chamber and laboratory studies, Final Report of the EC Project ENV4-CT97-0419, Mainz (2000). Tadiü, J., I. Juraniü and G.K. Moortgat; Photoxidation of n-hexanal in air, Molecules, 6 (2001) 287-299. Thiault, G., A. Mellouki, G. Le Bras and K. Wirtz; The photolysis of aromatic aldehydes in The European Photoreactor, EUPHORE, 4th report, (2001), 29. Thiault, G., A. Mellouki , G. Le Bras; Kinetics of gas phase reactions of OH and Cl with aromatic aldehydes; Phys. Chem. Chem. Phys, 4 (2002) 2194 - 2199. Thiault, G., A. Mellouki , G. Le Bras,A. Chakir, N. Sokolowski-Gomez, D. Daumont; UV-absorption cross sections of benzaldehyde, ortho-, meta-, and para-tolualdehyde, J. Photochem. Photobiol. A: Chem. 162 (2004) 273. Volkamer, R., U. Platt and K. Wirtz; OH reaction rate constants and photolysis frequencies of a series of aromatic aldehydes and phenols in The European Photoreactor, EUPHORE, 3rd report, (2000) 1. Wenger, J.C., S. Le Calvé, H.W. Sidebottom, K. Wirtz, M. Martín-Reviejo and J.A. Franklin; Photolysis of chloral under atmospheric conditions, Environ. Sci. Technol. 38 (2004) 831-837.

Determination of Photolysis Frequencies for Selected Carbonyl Compounds in the EUPHORE Chamber Romeo-Iulian Olariu1, Marius Duncianu1, Cecilia Arsene1, and Klaus Wirtz2 1

2

“Al.I. Cuza” University of Iasi, Faculty of Chemistry, Analytical Chemistry Department, 700506 Iasi, Romania. Centro de Estudios Ambientales del Mediterráneo, Parque Tecnológico, C.\ Charles R. Darwin 14, 46980 Paterna, Valencia, Spain Key Words: Photolysis, Actinic flux, Spectroradiometer, Carbonyl compounds

Introduction Small carbonyl compounds are formed during the photochemical oxidation of many volatile organic compounds (VOC’s), in urban as well as in rural areas. Photolysis and reaction with the OH radical are the most important initiation reactions for the atmospheric degradation of these compounds, and lead to the formation of peroxy radicals in the former case and either stable molecules and/or free radicals in the latter case (Finlayson-Pitts and Pitts, 1999). About 580 actinic flux spectra recorded under different meteorological conditions (clear sky, partially cloudy or overcast) in the EUPHORE smog chambers, have been used to calculate the photolysis frequencies for various small carbonyl compounds that are considered to be important from the atmospheric chemistry point of view. The results are presented here. Experimental a) Actinic Flux measurement (F) The actinic flux spectra were recorded under realistic atmospheric conditions in the large scale outdoor EUropean PHOtoREactor (EUPHORE). The EUPHORE chamber is made from PTFE-Teflon foil. It is half spherical in shape with a volume of about 204 m3 and is mounted on aluminium floor panels. The PTFE-Teflon foil is highly transparent even to short wavelength radiation, with transmissions ranging from 85% in the range 500 - 320 nm to > 75% at 290 nm. A detailed description of the experimental set-up is available elsewhere (Becker, 1996). The light intensity was measured using a calibrated spectroradiometer (Bentham DM300). Both direct and reflected beams of the incident light are optically coupled and pass simultaneously but geometrically separated through a double monochromator. The entire set of actinic flux spectra have been recorded in the range of 290 nm to 520 nm. The spectra were recorded every 5 min with a spectral resolution of 1 nm in the range from 290 nm to 360 nm and 5 nm between 360 to 520 nm. The actinic flux spectra were recorded during an entire year, i.e. under different meteorological conditions, clear sky, scattered clouds or overcast etc. b) Photolysis frequencies calculation (Jx) In practice, rather than integrating over the wavelength spectrum to obtain the total rate of photolysis, the sum of product V(O) × I(O) × F(O) over discrete wavelength intervals 'Oҏwas used. The choice of the width of these intervals was usually based on the intervals for which F(O) was measured. Thus, the total photolysis rate was calculated using the equation:

121 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 121–128. © 2006 Springer. Printed in the Netherlands.

122

R.-I. Olariu et al.

J

Oi

¦O

290

V (O ) u I O u F (O)

(E1)

where V (O ) and I (O ) are the values averaged over a wavelength interval 'O centred at O, and F(O) is actinic flux in photons cm-2 s-1 summed over the wavelength interval'Ocentred at O The numerical sum over the relevant wavelength range in eq. E1 has been performed using a home-produced MS-DOS program. The program contains a data base for the computation of 43 photo-dissociation reactions of 40 different trace gases of atmospheric relevance. For the present study the measured EUPHORE radiation spectra together with published molecular data (Moortgat, 2000) were used for the calculation of the photolysis frequencies of 17 carbonyl compounds.

c ) Calculation of solar zenith angle (T) The solar zenith angle can be calculated in the following manner for any particular location (i.e. latitude and longitude), day of the year and time of the day (Finlayson-Pitts and Pitts, 1999). There are two steps that need to be considered. First, the “local hour angle” (th), which is defined as the angle between the meridian of the observer and that of the sun, in radians, has to be calculated:

th = S[(GMT/12)-1 + (longitude/180)] + EQT

(E2)

where GMT is Greenwich mean time converted from the local time, longitude (in degree, west of the Greenwich meridian are negative) and EQT is the “equation time”, given by: -5

-3

EQT = 7.5 ×10 + 1.868 ×10 cos N -2

–3.2077 ×10 sin N -2

–1.4615 ×10 sin 2N - 4.0849 ×10 -2sin 3N

(E3)

where N (in radian) is defined as:

N= 2Sdn/365

(E4)

The day of the year, dn, is defined as the day number (0 - 364), with 0 corresponding to January 1 and 364 to December 31. The second parameter that is needed for the calculation of the solar zenith angle at a particular time and place is the solar declination, G, defined as the angle between the direction of the sun and the equatorial plain of the earth. The value of G (in radian) can be calculated as follows: -3

G = 6.918 ×10 -0.399912cos N + 0.070257 sin N – -3

-4

- 6.758 ×10 cos 2N + 9.07 ×10 sin 2N -3

-3

–2.697 ×10 cos 3N + 1.480 ×10 sin 3N

(E5)

Determination of Photolysis Frequencies for Selected Carbonyl Compounds

123

with these two parameters the solar zenith angle for any particular time and place can be easily calculated: cos T = sinG ×sin(latitude) + cos G×cos(latitude) ×cos th

(E6)

where Gand th are calculated as already described and latitudes north of the equator (in radian) are positive and south are negative. If all of the input parameters are in radians, T is also obtained in radians and can be converted to degree using 1 rad = 57.2960. For example, at Valencia, Spain, (longitude: 0.50 W, latitude: 39.50 N) on September 1 at noon local time, GMT = 11.0, dn= 243, N = 4.183, EQT = - 0.0016, th = -0.2721 rad, G = 0.149 rad, and cosT = 0.829, giving a solar zenith 0 angle of 0.592 rad, or 33.92 Consequently, the actinic flux data recorded in the EUPHORE chamber during one year will provide useful and complementary data over the range of the zenith angle values pertinent to this local area. Results and discussion The actinic flux spectra were used to calculate the spectral photolysis frequency distributions taking into account the molecular parameters (V(O) x I(O)), which after wavelength integration yield the respective J values. Figure 1 shows typical actinic flux profiles recorded in the EUPHORE chamber at various times of the year. The spectral data have been used to obtain atmospheric photo-dissociation frequencies for several carbonyl compounds. For all the investigated compounds (see Table 1) plots of the calculated photolysis frequencies versus solar zenith angle show a cloud-type distribution of data points. Figure 2 shows an example for acetaldehyde. The strong scattering and lack of a clear linear dependency has probably several causes one of which is likely the variation in the degree of cloudiness over the EUPHORE chamber. Since the photolysis frequency is directly coupled with the solar zenith angle one major objective of the present study was the realization of a relative dependency between the calculated photolysis frequencies and solar zenith angle. To obtain a clear dependency a statistical treatment of all the calculated photolysis frequencies, derived from all the actinic spectra recorded in the EUPHORE chamber has been performed. Figure 3 shows an example of the complete dataset for acetaldehyde. The statistical treatment has allowed a clear dependency between the calculated photolysis frequency and the solar zenith angle to be established. The result obtained for acetaldehyde statistical treatment after is presented graphically in Figure 4. The error bars represent the statistical (1V) error only. Photolysis frequencies have been calculated in the range of 19º to 71.5º solar zenith angles for 17 carbonyl compounds. The calculated photolysis frequencies obtained for the different zenith angles as derived from all the EUPHORE actinic flux spectra measurements are presented in Table 1. Using the data obtained in the present work a simple analytical expression has been derived for Jx(T) which captures more than 75% of all the measurements up to a 71.50 solar zenith angle with an error band of 14%. The general analytical expression obtained is as follows: Jx(T) = [m u exp(-n u sec(T)] s-1 2

E7

The empirical parameters m, n and r obtained from fits of the calculated photolysis frequencies as a function of solar zenith angle are listed in Table 2.

124

R.-I. Olariu et al.

Figure 1.

Actinic flux profiles recorded in the EUPHORE chamber for a 400 zenith angle at the following dates and times: a) on 24.03.2003 at 13:56 h; b) on 20.06.2002 at 16:00 h; c) on 20.03.2001 at 13:09 h; d) 20.03.2002 at 12:55 h.

Figure 2.

Plot of the calculated photolysis frequencies of acetaldehyde from the entire data set against solar zenith angle.

Determination of Photolysis Frequencies for Selected Carbonyl Compounds

Figure 3.

125

Calculated photolysis frequency profile for one year derived from data recorded using the EUPHORE outdoor chamber.

Photolysis frequenciey of acetaldehyde (s-1)

4.0e-5

3.0e-5

2.0e-5

1.0e-5

0.0 0

Figure 4.

10

20

30 40 50 60 Solar zenith angle (degree)

70

80

90

Plot of the calculated photolysis frequencies of acetaldehyde (obtained after statistical treatment) versus solar zenith angle.

R.-I. Olariu et al.

126 Table 1.

-1

Calculated photolysis frequencies for carbonyl compounds (in s ) at different zenith angles derived from the EUPHORE chamber measurements.

Compound

exp

acetaldehyde CH3CH2CHO n-Butyraldehyde i-Butyraldehyde 3-Methybutyraldehyde Pentanal Pinonalddehyde Pivalaldehyde Glyoxal Glycolaldehyde Acrolein Methacrolein Crotonaldehyde Caronaldehyde 2-Petanone MVK Pyruvic_acid

10 10-5 -5 10 -5 10 -5 10 10-5 10-5 -5 10 -3 10 -6 10 -4 10 -4 10 -4 10 10-5 10-5 -4 10 -4 10

-5

Table 2.

19.10 3.35±0.51 3.76 ±0.58 4.93±0.75 4.98±0.76 4.37±0.68 5.06±0.77 6.70±1.05 2.76±0.43 2.82±0.86 8.27±0.13 4.65±0.73 4.59±0.70 4.40±0.68 3.53±0.57 1.28±0.20 4.55±0.70 2.86±0.45

zenith angle (degree) 32.03 51.83 2.96±0.41 2.09±0.24 3.32±0.46 2.34±0.27 4.31±0.61 3.10±0.36 4.35±0.62 3.12±0.36 3.90±0.57 2.78±0.35 4.44±0.63 3.20±037 5.86±0.83 4.05±0.50 2.47±0.36 1.74±0.22 2.97±0.37 2.58±0.22 7.17±0.10 4.87±0.65 4.48±0.60 3.54±0.30 4.38±0.59 3.42±0.30 4.20±0.56 3.29±0.28 3.07±0.46 2.07±0.28 1.10±0.16 0.74±0.10 4.35±0.58 3.41±0.30 2.79±0.38 2.24±0.19

71.53 0.87±0.14 0.97±0.01 1.31±0.22 1.31±0.22 1.21±0.20 1.37±023 1.62±0.26 7.41±0.12 1.45±0.39 1.87±0.31 1.82±0.32 1.73±0.30 1.67±0.29 7.88±0.12 0.28±0.04 1.73±0.30 1.17±0.21

Empirical parameters m, n and r2 from E7.

2

compound

m

n

r

acetaldehyde CH3CH2CHO n-Butyraldehyde i-Butyraldehyde 3-Methybutyraldehyde Pentanal Pinonalddehyde Pivaldehyde Glyoxal Glycolaldehyde Acrolein Methacrolein Crotonaldehyde Caronaldehyde 2-Petanone MVK Pyruvic_acid

10-4 -4 10 -4 10 -4 10 -4 10 -4 10 -4 2 x 10 -4 10 -4 9 x 10 -4 10 -4 8 x 10 -4 8 x 10 -4 7 x 10 -4 10 -4 10 -4 8 x 10 -4 5 x 10

0.7291 0.760 0.737 0.741 0.758 0.736 0.807 0.779 0.356 0.869 0.484 0.509 0.503 0.889 0.884 0.503 0.457

0.96 0.98 0.98 0.98 0.97 0.98 0.98 0.97 0.98 0.98 0.98 0.98 0.98 0.97 0.98 0.99 0.99

Determination of Photolysis Frequencies for Selected Carbonyl Compounds

127

In Table 3 the mean of the all calculated photolysis frequencies for the various compounds obtained from the analysis of the actinic flux measurements performed using EUPHORE chamber facilities is compared with results from other similar studies. In the estimates a quantum yield of unity has been assumed. The errors are a combination of the statistical errors and the errors associated with uncertainties in the molecular parameters. The effective quantum yields, Ieff, in Table 3 were determined by division of the measured individual photolysis frequencies (J3) by the photolysis frequencies calculated in this study (J1) assuming a quantum yield of unity. For some of the carbonyl compounds the determined Ieff is significantly below unity. The photolytic lifetimes obtained using the present photolysis frequencies are also presented in Table 3. Table 3.

compound

Photolysis frequencies and effective quantum yield Ieff for some carbonyl compounds as derived from the EUPHORE facilities and found in the literature. exp -5

J1(X), s

-1

estimated this work

Acetaldehyde 10 2.26±0.84 -5 CH3CH2CHO 10 2.53±0.95 -5 3.34±1.23 n-Butyraldehyde 10 -5 3.37±1.25 i-Butyraldehyde 10 -5 2.97±1.19 3-Methyl-butyraldehyde 10 -5 3.44±1.27 Pentanal 10 -5 4.43±1.75 Pinonaldehyde 10 -5 1.86±0.71 Pivaldehyde 10 -3 2.54±0.47 Glyoxal 10 -6 8.52±1.58 Glycolaldehyde 10 -4 3.62±0.97 Acrolein 10 -4 Methacrolein 10 3.51±0.98 -4 Crotonaldehyde 10 3.37±0.93 -6 2-Pentanone 10 8.28±3.46 -4 MVK 10 3.49±0.97 -4 Pyruvic acid 10 2.26±0.58 a b Moortgat, 2000, Wirtz, 1999.

J2(X), s

-1

estimated literature a

4.9 a 3.8 a 5.1 a 5.2 a 4.7 a 5.4 a 8.0 a 2.6 a 2.7 a 8.6 a 4.3 a 5.2 a 4.0 b 8.6 a 5.2 a 2.3

J3(X), s

-1

Ieff

measured values a

0.29±0.27 a 1.10±0.10 a 1.00±0.20 a 3.70±0.10 a 1.25±0.10 a 1.60±0.20 a 1.15±0.10 a 1.45±0.10 a 0.10±0.03 a 11.4±2.5 a <0.01 a <0.01 a 0.11±0.01 b 0.6 a <0.01 a 1.00±0.15

0.12 0.46 0.29 1.09 0.42 0.46 0.26 0.78 0.039 1.33 <0.003 <0.003 0.03 0.07 <0.003 0.44

Remarks In the present work a large set of actinic spectra recorded in the EUPHORE chamber under various atmospheric conditions has been obtained and used for the calculation of photolysis frequencies of 17 organic carbonyl compounds. From a statistical analysis of the photolysis frequencies calculated for the compounds an analytical form for JX(T) has been derived. For unsaturated compounds (methyl vinyl ketone, methacrolein, acrolein and crotonaldehyde) Ieff is negligible, although those cxompounds possess absorption spectra reaching the near visible.

128

R.-I. Olariu et al.

In atmospheric models currently used, quantum yields are often assumed to be unity, which is seldom the case, as is shown in this study. This leads to overestimation of the calculated photodissociation rates and the associated radicals formed in the photolysis processes. These new data will reduce the present uncertainties associated with photolysis processes, so that future assessments can be made with more confidence. Acknowledgments Financial supports of this work from the European Commission within the 5th Framework Programme project IALSI and action travel funding from Joint Research Centre, Ispra, Italy, to NATO EST. ARW 980164 Research Workshop are gratefully acknowledged. References Finlayson-Pitts, B. J. and J. N. Pitts. Jr.; Chemistry and Physics of the Upper and Lower Atmosphere, Wiley, New York, 1999. Becker, K. H.; The European Photoreactor “EUPHORE”, Final report , EC contract EV5V-CT92-0059, 1996. Moortgat, G. K.; Evaluation of Radical Sources in Atmospheric Chemistry trough Chamber and Laboratory Studies, “RADICAL”, Final Report on EU Project, EC contract ENV4-CT97-0419, 2000. Wirtz K.; lecture presented at the combined US/Germany Environmental Chamber Workshop, Riverside, 1999.

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy (DOAS): Advancements in Simulation Chamber and Field Experiments R. Volkamer1, I. Barnes2, U. Platt3, L. T. Molina1, and M. J. Molina1 1

Department of Earth, Atmosphere and Planetary Sciences, MIT, 77, Massachusetts Av. 54-1320, Cambridge, MA 02139, USA 2 Bergische Universität Wuppertal, FB C - Physikalische Chemie, Gauss Strasse 20, 42119 Wuppertal, Germany 3 Institut für Umweltphysik, University of Heidelberg, D-69120 Heidelberg, Germany Key Words: Glyoxal, DOAS, Atmosphere, Sources, Photolysis, Isoprene

Introduction Air pollution in many large cities is linked with the photochemical transformation of primary pollutants like VOCs (volatile organic compounds) and NOx, which in the presence of sunlight foster the formation of secondary pollutants including ozone (O3) and secondary organic aerosol (SOA) (Finlayson-Pitts and Pitts 2000; Molina and Molina 2002). ‘Photochemical smog’ has adverse effects on human health (Kunzli et al. 2000; Evans et al. 2002), the ecosystem (Middleton et al. 1950; Gregg et al. 2003) and regional climate (Lelieveld et al. 2001; Ramanathan and Crutzen 2003). While the crucial role of the atmospheric chemistry of VOCs in the formation of urban smog is well established, the performance of traditional indicators for VOC chemistry like O3, Ox (sum of O3 and nitrogen dioxide (NO2)), formaldehyde (HCHO), and methylvinylketone (MVK) is affected by direct vehicle emissions, which contribute significantly to the atmospheric concentrations of these species; in the case of O3, high concentrations of nitric oxide (NO) efficiently scavenge this gaseous pollutant during morning rush hour (Finlayson-Pitts and Pitts 2000). Hence, these trace gases may be ambiguous indicators for VOC chemistry in urban air, where the simultaneous presence of multiple pollutant sources complicates the development of robust control strategies to reduce O3 and aerosol levels. Recently, glyoxal (CHOCHO) was detected for the first time directly in the atmosphere by means of differential optical absorption spectroscopy (DOAS) (Volkamer et al. 2005a). These time-resolved glyoxal measurements revealed atmospheric glyoxal concentrations to be essentially unaffected by vehicle emissions in Mexico City, and demonstrated that glyoxal is a unique indicator of fast VOC photochemical oxidation in urban air. Glyoxal is the smallest Į-dicarbonyl hydrocarbon and a mutagenic oxidation product (Kasai et al. 1998) from the oxidation of numerous VOCs (Calvert et al. 2000; Calvert et al. 2002). In a recent study using DOAS at the European Photo Reactor (EUPHORE) the glyoxal yield from the OH-radical initiated oxidation of benzene, toluene and p-xylene have been investigated (Volkamer et al. 2001). These time-resolved glyoxal yields allowed an improvement

129 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 129–141. © 2006 Springer. Printed in the Netherlands.

130

R. Volkamer et al.

in the chemical codes of aromatic VOC oxidation (Bloss et al. 2004b). However, quantitative mechanistic conclusions were limited by the then uncertain DOAS calibration, namely the UV/vis absorption cross-section of glyoxal. An improved UV/vis absorption cross-section of glyoxal has since been recorded and now allows this limitation to be overcome (Volkamer et al. 2005b) and has enabled further improvements in mechanistic conclusions about aromatic oxidation. During the day, rapid photolysis of glyoxal is a source of radicals. Glyoxal loss due to photolysis and reaction with OH-radicals limit its atmospheric lifetime during the day (Plum et al. 1983; Atkinson 2000). The atmospheric lifetime with respect to these gas-phase glyoxal sinks was recently quantified experimentally in Mexico City and found to be considerably shorter than previously believed, i.e. about 1-2 hours during most of the day (Volkamer et al. 2005a). On ‘cloudless’ days, glyoxal photolysis accounts for about two thirds of the gas-phase glyoxal loss; OH-radical reactions are relatively more important in the presence of clouds, at elevated solar zenith angles (SZA>80°), and at night. The SZA dependence of glyoxal photolysis is discussed in more detail in this contribution. In this contribution the re-evaluated yields from the OH-radical initiated oxidation of benzene, toluene, p-xylene, and initial results of new simulation chamber experiments on ‘prompt’ glyoxal formation from isoprene oxidation are presented. A detailed discussion of sources, sinks and their uncertainties to model atmospheric concentrations of glyoxal is presented, and exemplifies how basic research in environmental simulation chambers besides giving input for photochemical models also triggers advancements with measurement techniques for field observations. The integration of laboratory and field observations by models in turn will guide future research on atmospheric chemical processes. Experimental During April 2003, a field campaign was held to investigate the formation of photochemical smog in the Mexico City Metropolitan Area (MCMA). Long-Path DOAS (LPDOAS) was used to detect glyoxal using its unique specific narrow-band (< 5 nm) absorption structures in the visible spectral range, that enable DOAS (Platt 1994) to separate trace gas absorptions from broadband molecule and aerosol extinction in the open atmosphere. As part of the MCMA-2003 field campaign, two LP-DOAS instruments were deployed at the CENICA super-site. Only data from DOAS2 are presented here, which faced an array of retroreflectors at Museo Fuego Nuevo, Cerro de la Estrella. The total atmospheric path length was 4420m at a mean height of 70m above ground level. Atmospheric spectra were recorded at a spectral resolution of 0.4 nm Full Width at Half Maximum (FWHM) about every 2 to 15 minutes. Glyoxal was retrieved in the spectral range between 420 and 465 nm, covering three prominent glyoxal absorption lines using a 5th order polynomial high pass filter and non-linear least squares fitting of glyoxal, NO2, O4, H2O, and two lamp-reference spectra using the Windoas evaluation software (Fayt and van Roozendael 2001). As reference spectra available high-resolution absorption cross-section spectra (see (Alicke et al. 2003) and references therein) were degraded to match the instrument resolution. The DOAS detection of glyoxal is inherently calibrated from

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

131

knowledge of its differential absorption cross-section (Volkamer et al. 2005b), which for the three covered absorptions lines were 5.5, 2.5, and 1.7×10-19 cm2 at 455, 440, and 428 nm, respectively. Absorption features with differential optical densities of 8×10-4 could be detected, corresponding to a mean detection limit of 3×109 molec/cm3 or 0.15 ppbv (1 ppbv = 1 parts per billion by volume = 1000 pptv § 1.92 ×1010 molec/cm3 in Mexico City). The formation of glyoxal in the early stage of the OH-radical initiated oxidation of isoprene (‘prompt’ glyoxal) in the presence of NOx was also investigated using DOAS at the EUPHORE chamber. A general description of EUPHORE is available in the literature (Klotz et al. 1998). A detailed description of the DOAS instrument at EUPHORE and the relative yield approach used here is available in Volkamer et al. (2001) and Volkamer (2001). In brief, measured product concentrations are corrected for losses through reaction with OH-radicals, photolysis and wall-deposition. The ratio of the yield of a target product (ɎTAR) relative to the known yield of a reference product (ɎREF) is determined from: ɎTAR / ɎREF = [TAR]corr / [REF] corr From a plot of the corrected concentration of the target product as a function of the corrected concentration of the reference product a straight line is expected if both products are formed as primary products. Any secondary formation of the product leads to deviations from the linear relationship. The yield ratio is determined from the slope of the data. The relative yield method is employed in this work to determine the primary glyoxal yield relative to methacrolein, which is an established primary product from the OH-radical initiated oxidation of isoprene. For the discussion of glyoxal sources from VOC oxidation, glyoxal yields were calculated for VOCs where no measurements are reported in the literature. Suitable single VOC compound subsets of the Master Chemical Mechanism MCMv3.1 (Saunders et al. 2003; Bloss et al. 2004a; Bloss et al. 2004b) were prepared. The code was initialised using the following initial conditions: [VOC]i=50ppbv, SO2=1ppbv, CO=120ppbv, CH4=1800ppbv, O3=50ppbv, H2O2=2ppbv, NO2=7ppbv, variable NO (1-100 ppbv) and mid-latitude conditions, and run for three consecutive days. Glyoxal yields were derived as the overall moles of glyoxal formed (loss corrected) divided by the moles of VOCi reacted after one day. Results and Discussion As part of this MCMA-2003 campaign, glyoxal was detected for the first time directly in the atmosphere (Volkamer et al. 2005a). As is visible from the concentration time profile shown in Figure 1, the rapid variation of glyoxal was well resolved by these time resolved DOAS measurements, which present useful data to validate the performance of photochemical models that are used as tools for the design of control strategies to improve air quality in Mexico City.

132

Figure 1.

R. Volkamer et al.

Concentration time profile of glyoxal measured on 10 April 2003 in Mexico City. Letters indicate times of the corresponding atmospheric spectrum shown on the right. Spectra have been vertically shifted for clarity; thick lines indicate scaled reference spectra of glyoxal; error bars indicate 2-sigma level. Reproduced from (Volkamer et al. 2005a).

Previous indirect measurements of glyoxal by cartridge collection/chromatography methods do not resolve the rapidly changing temporal variation of glyoxal in urban air. These are highly specialized and time intensive techniques (Grosjean et al. 1999; Moortgat et al. 2002), which involve pre-concentration of glyoxal and many other carbonyls on a solid sorbent (cartridge) covered with either DNPH (2,4-dinitrophenylhydrazine) or PFBHA (O-(2,3,4,5,6pentafluorobenzyl) hydroxylamine) derivatisation agent, followed by solvent extraction and gas/liquid chromatography separation and mass spectrometric/UV-visible spectroscopic detection. Possible complications may arise if no differentiation between hydroxyacetaldehyde and glyoxal can be made (Grosjean et al. 1996a). The time resolution with such measurements depends on sample accumulation time (one to four hours) and detection limits depend on sample volume and typically range between 50 - 200 ppt. The detection of glyoxal by DOAS in the open atmosphere poses several advantages, including the absence of sampling artefacts, selective detection, inherent calibration, and the combination of high time-resolution (few minutes) with reasonable sensitivity (on the order of 100 ppt). Atmospheric glyoxal concentrations measured by DOAS and other methods are listed in Table 1. Glyoxal levels in the MCMA are well within the concentration range observed in other

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

133

urban environments, and indeed do not seem exceptionally high, suggesting that routine DOAS measurements of glyoxal are feasible also in other locations. Table 1.

Atmospheric concentrations of glyoxal.

Reference

Location

Airmass

(Volkamer et al. 2005a) (Ho and Yu 2002) (Kawamura et al. 2000) (Grosjean et al. 1996b)

Mexico City, Mexico Hong Kong, China Los Angeles, USA Los Angeles mean Long beach Central LA Azusa Claremont Las Vegas, USA Summer Winter Porto Allegre, Brazil Giesta, Portugal Pabsthum, Germany Georgia, USA

polluted urban polluted urban polluted urban polluted urban

(Jing et al. 2001)

(Grosjean et al. 1999) (Borrego et al. 2000) (Moortgat et al. 2002) (Lee et al. 1995)

Mixing ratio [ppb] < 0.15 – 1.82 0.5 – 4.1 0.04 – 0.95 0.78 ± 0.85 <0.12 – 0.8 <0.12 – 2.2 <0.12 – 3.0 0.4 – 3.8

urban

urban semi-urban Semi-rural rural

0.12 – 0.42 0.09 – 0.21 0.3 0.52 – 2.42 0.01 – 0.12 0.02 – 0.19

A compilation of glyoxal sources from VOC oxidation is presented in Table 2. Experimental data on the glyoxal branching ratio is given where available; average values are listed if several values are available (Calvert et al. 2000; Calvert et al. 2002). For aromatic VOC elevated concentrations of NO and NO2 were recently demonstrated to affect the relative importance of the various oxidation routes (Klotz et al. 2002; Volkamer et al. 2002; Atkinson and Arey 2003). This influence is also visible in the glyoxal yields (see e.g. Table 5.8 of Volkamer (2001)), and values in Table 2 reflect yields measured under close to atmospheric NOx concentrations. For VOCs marked with an asterix, average yields of glyoxal were calculated using the MCMv3.1 code as described above. As is apparent from Table 2, aromatic VOC show the highest glyoxal yields. The uncertainty of the re-evaluated yields (Volkamer et al. 2001; Volkamer et al. 2005b) is reduced by 50% and no longer limits mechanistic conclusions. Assuming a glyoxal/methylglyoxal ratio of 3:2 from toluene (Smith et al. 1998), and a nitrate yield from the bicycloalkylperoxi radical + NO reaction of 10% (Jenkin et al. 2003; Saunders et al. 2003), this glyoxal yield corresponds to a branching ratio of the bicycloalkyl radical pathway from benzene and toluene of 32% and 57.5%. Including quantified ring-retaining product yields (Klotz et al. 1998; Smith et al. 1998; Volkamer et al. 2002) about 85±8% and 82±6% of the primary oxidation routes are identified (molar basis, errors indicate 2-sigma uncertainty level). This is significantly below unity, indicating that a currently unassigned reaction pathway is needed in order to close the carbon balance in the early stages of the oxidation of benzene and toluene.

134 Table 2.

R. Volkamer et al. Sources of glyoxal from the oxidation of VOCs.

Glyoxal precursor VOC

Oxidant

Glyoxal yield

Reference

benzene OH 32 ± 5 (Volkamer et al. 2005b) benzene* OH 36 estimate this work benzene** OH 53 - 70 estimate this work toluene OH 30.6 ± 6 (Volkamer et al. 2005b) p-xylene OH 31.9 ± 5 (Volkamer et al. 2005b) m-xylene OH 7.9 ± 0.4 (Bandow and Washida 1985a) o-xylene OH 8±4 (Bandow and Washida 1985a) 1,2,4-trimethylbenzene OH 7.8 ± 0.5 (Bandow and Washida 1985b) 1,2,3-trimethylbenzene OH 7.2 ± 0.1 (Bandow and Washida 1985b) ethylbenzene** OH 52 – 58 estimate this work n-propylbenzene** OH 54 – 56 estimate this work i-propylbenzene** OH 55 – 60 estimate this work o-ethyltoluene** OH 14 – 18 estimate this work m-ethyltoluene** OH 2.5 – 4.7 estimate this work p-ethyltoluene** OH 39 – 42 estimate this work acetylene* OH 63.5 estimate this work 2-methyl 1,3-butadiene (isoprene) OH 3 this worka 2-methyl 1,3-butadiene (isoprene)** OH/O3 7–8 estimate this workb 2-methyl 1,3-butadiene (isoprene)** OH/O3 ”2 (Paulson and Seinfeld 1992) ȕ-pinene** OH/O3 4.5 – 5.5 estimate this work ethene O3 0.44 (Calvert et al. 2000) ethene** OH/O3 < 1 – 3.6 estimate this work propene O3 8.3 ± 0.6 (Calvert et al. 2000) 1-butene O3 0.3 ± 0.5 (Calvert et al. 2000) cis-2-butene O3 11.3 ± 5.9 (Calvert et al. 2000) trans-2-butene O3 14.7 ± 6.2 (Calvert et al. 2000) 2-methyl-2-butene O3 4.5 ± 2.1 (Calvert et al. 2000) 2-methylpropene O3 0.5 ± 0.4 (Calvert et al. 2000) 1,3-butadiene O3 0.4 ± 0.3 (Calvert et al. 2000) 1-pentene O3 2±2 (Calvert et al. 2000) 2-methyl-1-butene O3 0.4 ± 0.8 (Calvert et al. 2000) 2-methyl-2-butene O3 4.5 ± 2.1 (Calvert et al. 2000) cis-4-octene O3 5.8 ± 0.3 (Calvert et al. 2000) trans-4-octene O3 3 ± 0.2 (Calvert et al. 2000) trans-3,4-dimethyl-3-hexene O3 11.8 ± 1.6 (Calvert et al. 2000) 3,4-diethyl-2-hexene O3 8.5 ± 1.2 (Calvert et al. 2000) Cyclohexene O3 15 ± 1 (Calvert et al. 2000) n-butane** OH < 0.1 estimate this work 2,2-dimethyl butane** OH < 0.2 estimate this work 3-methyl hexane** OH < 0.003 estimate this work 2-methylpentane** OH < 0.5 estimate this work acrolein OH 7.5 ± 2.5 (Magneron et al. 2002) trans-crotonaldehyde OH 16 ± 4 (Magneron et al. 2002) ethyl 1-propenyl ether O3 16.7 ± 0.2 (Grosjean and Grosjean 1999) 4-hexene-3-one O3 6.4 ± 0.3 (Grosjean and Grosjean 1999) cis-3-hexenyl acetate O3 0.3 ± 0.1 (Grosjean and Grosjean 1999) *calculated primary glyoxal yield; **calculated overall glyoxal formation (primary and secondary oxidation product); a experimental value for the primary glyoxal yield from the isoprene + OH reaction; b not including primary glyoxal formation (see a).

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

135

Despite the significant progress which has been made in recent years on the understanding of aromatic oxidation routes, care must be taken with predictions of glyoxal from chemical codes. Beyond the first generation of oxidation products very little information is as yet available about the identity and fate of the products following ring-cleavage, even for the most abundant alkylbenzenes (Bloss et al. 2004a; Bloss et al. 2004b). Oxidation routes are highly speculative as can be exemplified for benzene, where the primary glyoxal yield is identical to the overall glyoxal formation after several hours of oxidation (Volkamer et al. 2001). This is only partly reproduced by the MCM code, which reflects the primary glyoxal yields but overestimates glyoxal formation in the later stage of the experiment. The predicted overall glyoxal yields from benzene range between 53% and 70%, depending on the initial conditions. Higher yields are predicted at lower initial NOx concentrations. This apparently significant secondary formation of glyoxal from benzene is not in agreement with the experimental observations, as is the apparently reversed NOx dependence of the yields. In analogy to toluene and p-xylene, glyoxal yields from benzene are expected to be lower in the absence of NOx (see e.g. Table 5.8 of Volkamer (2001)). While the formation of glyoxal from aromatic VOC is likely over-predicted by the MCM code (in part due to the lower re-evaluated glyoxal yields), glyoxal formation is under estimated from alkenes like cis- and trans-2-butene, for which glyoxal is not a predicted oxidation product. Glyoxal is formed with high yield from acetylene, which will likely be only a minor source in urban areas, and relatively more important in remote air, when more reactive VOCs are depleted. The glyoxal source from alkanes and oxygenates is expected to be minor as a consequence of low glyoxal yields and/or VOC concentrations.

Figure 2.

Glyoxal and methacrolein formation from the OH-radical initiated oxidation of isoprene during the experiment ISO2 at EUPHORE (27.03.02).

136

R. Volkamer et al.

Glyoxal is formed from the oxidation of biogenic VOC like isoprene and ȕ-pinene in nonnegligible amounts Yields are, however, currently not well established. Existing isoprene oxidation schemes (Paulson and Seinfeld 1992; Saunders et al. 2003; Geiger et al. 2003) rationalize glyoxal formation exclusively via the further oxidation of stable intermediate compounds (as secondary glyoxal). Recent quantum calculations have predicted isoprene to form glyoxal as a primary oxidation product (Dibble, 2002). The proposed reaction mechanism involves rapid E/Z isomerization of excited OH-isoprene adducts, followed by 1,5-H-shift of the E-isomer and subsequent glyoxal formation. The hypothesis of primary glyoxal formation the oxidation of isoprene was tested in an experiment at EUPHORE, where glyoxal and methacrolein were measured at high timeresolution by DOAS. Figure 3 shows the data from the first 50mins of reaction time. Relative to the established primary oxidation product methacrolein, glyoxal concentrations increase in a fixed relationship, indicating that glyoxal is indeed formed as a primary oxidation product in the OH-radical initiated oxidation of isoprene. The proposed novel chemistry would appear to be operative. A value of 0.014 has been determined for the ratio of the primary glyoxal and methacrolein yields. Assuming a methacrolein yield from isoprene of 21%, the primary glyoxal yield is about 3%. This novel chemistry needs be incorporated into isoprene oxidation schemes, and will lead to higher predicted glyoxal concentrations in areas with active isoprene chemistry. In Figure 3 the effect of spectral resolution on the absorption features of glyoxal is shown in panel A. The high-resolution absorption cross-section (Volkamer et al. 2005b) was convoluted to lower resolution, using a Gaussian shape line function of 1nm full width half maximum (see inset of panel A). Glyoxal photolysis as a function of wavelength was calculated as described in (Volkamer et al. 2005b) for solar zenith angles of 0, 60 and 80 degrees. In panels B and C, data for overhead sun are drawn as solid lines (dark shaded area in panel C, SZA = 0 degrees), 60 degrees as dashed lines (light grey shaded area), 80 degrees as dotted lines (meshed area), and integrated photolysis frequencies (J-values) correspond to J0=1.17×10-4, J60=5.33×10-5 and J80=1.22×10-5 s-1. While the J0/J60-ratio matches the ratio of photon actinic fluxes of visible light (near 415nm) within 25-30%, differences become larger at higher SZA, i.e. the J60/J80 ratio deviates from the ratio of visible photon actinic fluxes by more than a factor of ten, reflecting that most glyoxal photolysis takes place at shorter wavelengths. In panel C, the photolysis rate is shown accumulated over wavelength (right scale). About one third to half of glyoxal photolysis takes place at wavelengths below 335 nm, where photolysis quantum yields are relatively well known. Current best estimates of quantum yields (RADICAL 2001) suggest that about half to two thirds of glyoxal photolysis takes place at wavelengths > 335 nm, where quantum yields are considerable more uncertain. Despite recent progress in our understanding of glyoxal photolysis quantum yields, most glyoxal photolysis is taking place at wavelengths where quantum yields are currently not well established. A better understanding of the wavelength (and pressure) dependence of quantum yields is desirable.

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

Figure 3.

137

Glyoxal absorption cross-section and photolysis frequencies (J-values). (A) High resolution absorption cross-section of glyoxal (dotted line) is convoluted to 1nm resolution (solid line) by integrating with a Gaussian shape line function of about 1nm width (inset). (B) photon actinic flux for overhead sun (SZA = O˚ , solid ဨ line), and SZA = 60˚ (dashed line) and SZA = 80˚ (dotted line) and photolysis quantum yield (grey shaded). (C) Glyoxal photolysis as a function of wavelength for SZA = O˚, 60˚ and 80˚ (shaded areas) and accumulated glyoxal photolysis (lines).

Conclusions Measurements of glyoxal (CHOCHO) by Differential Optical Absorption Spectroscopy (DOAS) are feasible on a routine basis in the Mexico City Metropolitan Area (MCMA) and other

138

R. Volkamer et al.

urban environments. The DOAS measurement of glyoxal combines the advantage of real-time detection in the open atmosphere, absence of sampling losses, high time-resolution, reasonable sensitivity, and inherent calibration from the knowledge of the absorption cross-section. Available high-resolution spectra enable the routine detection of glyoxal in most urban air masses and possibly also rural locations. Glyoxal is formed from the oxidation of numerous VOCs. Current modelling capabilities to predict the sources of glyoxal from VOC oxidation require revision of VOC chemical codes. In urban air, where ozone concentrations tend to be low, aromatic VOC chemistry is the dominant glyoxal source and glyoxal formation is likely over estimated as a result of significant, though experimentally not verified, secondary glyoxal from the MCMv3.1 code. In rural areas, the prediction of glyoxal sources is highly uncertain due to the unknown NOx dependence of glyoxal yields. Also, proposed reaction pathways in the isoprene + OH reaction, which lead to the formation of glyoxal as a primary product, have been experimentally verified. While the primary glyoxal yield is low (about 3%), this novel chemistry may be a potentially important glyoxal source in areas of high isoprene emissions. Current isoprene oxidation schemes do not reflect this novel chemistry and likely underestimate glyoxal production. Uncertainties in the sources of glyoxal from VOC chemistry may be hard to detect from model predictions that do not adequately represent the faster than expected glyoxal photolysis. Quantification of gas-phase sink reactions of glyoxal would benefit from a re-determination of the OH-radical rate constant of glyoxal, knowledge of the NO3-radical rate constant as well as a better understanding of the wavelength dependent photolysis quantum yields. Additional glyoxal losses from dry deposition and transfer to aerosols are expected (Jang et al. 2002; Kalberer et al. 2004), but too little information is as yet available to assess the relevance of these sinks for atmospheric residence times. In order to accurately predict glyoxal concentrations in the atmosphere, specific laboratory experiments to improve chemical codes with respect to sources and sinks of glyoxal are needed. Glyoxal measurements present a very useful indicator molecule for volatile organic compound (VOC) chemistry, and provide novel means to test predictions of photochemical models in urban and possibly rural environments. Clearly, the feasibility of routine DOAS measurements of glyoxal opens the possibility to advance our understanding of atmospheric chemical processes in urban and remote environments, and reproducing atmospheric observations presents a challenge for the future. Acknowledgements Financial support from the German BMBF (Program on Atmospheric Research AFO2000), US-National Science Foundation (NSF) and Mexican Metropolitan Environmental Commission (CAM) is gratefully acknowledged. R.V. gratefully acknowledges successive fellowships by the Camille and Henry Dreyfus Foundation and the Alexander von Humboldt Foundation.

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

139

References Alicke, B., A. Geyer, A. Hofzumahaus, F. Holland, S. Konrad, H.W. Patz, J. Schafer, J. Stutz, A. Volz-Thomas, and U. Platt; OH formation by HONO photolysis during the BERLIOZ experiment, Journal of Geophysical Research-Atmospheres, 108 (2003) 8247-8263. Atkinson, R.; Atmospheric chemistry of VOCs and NOx, Atmospheric Environment, 34 (2000) 2063-2101. Atkinson, R. and J. Arey; Atmospheric Degradation of Volatile Organic Compounds, Chemical Review, 103 (2003) 4605-4638. Bandow, H. and N. Washida; Ring-Cleavage Reactions of Aromatic-Hydrocarbons Studied by Ft-Ir Spectroscopy .2. Photooxidation of Ortho-Xylenes, Meta-Xylenes, and Para-Xylenes in the Nox-Air System, Bulletin of the Chemical Society of Japan, 58 (1985a) 2541-2548. Bandow, H. and N. Washida; Ring-Cleavage Reactions of Aromatic-Hydrocarbons Studied by Ft-Ir Spectroscopy .3. Photooxidation of 1,2,3-Trimethylbenzenes, 1,2,4-Trimethylbenzenes, and 1,3,5-Trimethylbenzenes in the Nox-Air System, Bulletin of the Chemical Society of Japan, 58 (1985b) 2549-2555. Bloss, C., V. Wagner, A. Bonzanini, M.E. Jenkin, K. Wirtz, M. Martin-Reviejo, and M.J. Pilling; Evaluation of detailed aromatic mechanisms (MCMv3 and MCMv3.1) against environmental chamber data, Atmospheric Chemistry and Physics Discussion, 4 (2004a) 5683-5731. Bloss, C., V. Wagner, M.E. Jenkin, R. Volkamer, W.J. Bloss, J.D. Lee, D.E. Heard, K. Wirtz, M. Martin-Reviejo, G. Rea, J.C. Wenger, and M.J. Pilling; Development of a detailed chemical mechanism (MCMv3.1) for the atmosheric oxidation of aromatic hydrocarbons, Atmospheric Chemistry and Physics Discussions, 4 (2004b) 5733-5788. Borrego, C., P. Gomes, N. Barros, and A.I. Miranda; Importance of handling organic atmospheric pollutants for assessing air quality, Journal of Chromatography A, 889 (2000) 271-279. Calvert, J.G., R. Atkinson, K.H. Becker, R.H. Kamens, J.H. Seinfeld, T.J. Wallington, and G. Yarwood (Eds.) (2002), The mechanisms of atmospheric oxidation of aromatic hydrocarbons, Oxford University Press, Oxford. Calvert, J.G., R. Atkinson, J.A. Kerr, S. Madronich, G.K. Moortgat, T.J. Wallington, and G. Yarwood (Eds.) (2000), The Mechanisms of Atmospheric Oxidation of the Alkenes, Oxford University Press, New York. Dibble, T.S.; Isomerization of OH-isoprene adducts and hydroxyalkoxy isoprene radicals, Journal of Physical Chemistry A, 106 (2002) 6643-6650. DOI: 10.1021/jp025682m Evans, J., J. Levy, J. Hammit, C. Santos Burgoa, and M. Castillejos (2002), Health Benefits of Air Pollution Control, in Air Quality in the Mexico Megacity, edited by Molina, M.J. and L.T. Molina, chapter 4, 105-136, Kluwer Academic Publishers, Dordrecht, The Netherlands. Fayt, C. and M. van Roozendael; WinDoas 2.1 – Software User Manual, (2001) Finlayson-Pitts, B. and J.N. Pitts (Eds.) (2000), Chemistry of the Upper and Lower Atmosphere, Academic Press, San Francisco. Geiger, H., I. Barnes, J. Bejan, T. Benter, and M. Spittler; The tropospheric degradation of isoprene: an updated module for the regional atmospheric chemistry mechanism, Atmospheric Environment, 37 (2003) 15031519. Gregg, J.W., C.G. Jones, and T.E. Dawson; Urbanization effects on tree growth in the vicinity of New York City, Nature, 424 (2003) 183-187. Grosjean, E., J.B. DeAndrade, and D. Grosjean; Carbonyl products of the gas-phase reaction of ozone with simple alkenes, Environmental Science & Technology, 30 (1996a) 975-983.

140

R. Volkamer et al.

Grosjean, E., P.G. Green, and D. Grosjean; Liquid chromatography analysis of carbonyl (2,4dinitrophenyl)hydrazones with detection by diode array ultraviolet spectroscopy and by atmospheric pressure negative chemical ionization mass spectrometry, Analytical Chemistry, 71 (1999) 1851-1861. Grosjean, E. and D. Grosjean; The Reaction of Unsaturated Aliphatic Oxygenates with Ozone, Journal of Atmospheric Chemistry, 32 (1999) 205-232. Grosjean, E., D. Grosjean, M.P. Fraser, and G.R. Cass; Air quality model evaluation data for organics .2. C-1-C-14 carbonyls in Los Angeles air, Environmental Science & Technology, 30 (1996b) 2687-2703. Ho, S.S.H. and J.Z. Yu; Feasibility of collection and analysis of airborne carbonyls by on-sorbent derivatization and thermal desorption, Analytical Chemistry, 74 (2002) 1232-1240. Jang, M.S., N.M. Czoschke, S. Lee, and R.M. Kamens; Heterogeneous Atmospheric Aerosol Production by AcidCatalyzed Particle-Phase Reactions, Science, 298 (2002) 814-817. Jenkin, M.E., S.M. Saunders, V. Wagner, and M.J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3 (Part B): tropospheric degradation of aromatic volatile organic compounds, Atmospheric Chemistry and Physics, 3 (2003) 181-193. Jing, L.H., S.M. Steinberg, and B.J. Johnson; Aldehyde and monocyclic aromatic hydrocarbon mixing ratios at an urban site in Las Vegas, Nevada, Journal of the Air & Waste Management Association, 51 (2001) 13591366. Kalberer, M., D. Paulsen, M. Sax, M. Steinbacher, J. Dommen, A.S.H. Prevot, R. Fisseha, E. Weingartner, V. Frankevich, R. Zenobi, and U. Baltensperger; Identification of polymers as major components of atmospheric organic aerosols, Science, 303 (2004) 1659-1662. Kasai, H., N. Iwamoto-Tanaka, and S. Fukada; DNA modifications by the mutagen glyoxal: adduction to G and C, deamination of C and GC and GA cross-linking, Carcinogenesis, 19 (1998) 1459-1465. Kawamura, K., S. Steinberg, and I.R. Kaplan; Homologous series of C-1-C-10 monocarboxylic acids and C-1-C-6 carbonyls in Los Angeles air and motor vehicle exhausts, Atmospheric Environment, 34 (2000) 4175-4191. Klotz, B., S. Sorensen, I. Barnes, K.H. Becker, T. Etzkorn, R. Volkamer, U. Platt, K. Wirtz, and M. Martin-Reviejo; Atmospheric oxidation of toluene in a large-volume outdoor photoreactor: In situ determination of ringretaining product yields, Journal of Physical Chemistry A, 102 (1998) 10289-10299. Klotz, B., R. Volkamer, M.D. Hurley, M.P.S. Andersen, O.J. Nielsen, I. Barnes, T. Imamura, K. Wirtz, K.H. Becker, U. Platt, T.J. Wallington, and N. Washida; OH-initiated oxidation of benzene - Part II. Influence of elevated NOx concentrations, Physical Chemistry Chemical Physics, 4 (2002) 4399-4411. Kunzli, N., R. Kaiser, S. Medina, M. Studnicka, O. Chanel, P. Filliger, M. Herry, F. Horak, V. Puybonnieux-Texier, P. Quenel, J. Schneider, R. Seethaler, J.C. Vergnaud, and H. Sommer; Public-health impact of outdoor and traffic-related air pollution: a European assessment, Lancet, 356 (2000) 795-801. Lee, Y.N., X.L. Zhou, and K. Hallock; Atmospheric carbonyl compounds at a rural southeastern United States site, Journal of Geophysical Research-Atmospheres, 100 (1995) 25933-25944. Lelieveld, J., P.J. Crutzen, V. Ramanathan, M.O. Andreae, C.A.M. Brenninkmeijer, T. Campos, G.R. Cass, R.R. Dickerson, H. Fischer, J.A. de Gouw, A. Hansel, A. Jefferson, D. Kley, A.T.J. de Laat, S. Lal, M.G. Lawrence, J.M. Lobert, O.L. Mayol-Bracero, A.P. Mitra, T. Novakov, S.J. Oltmans, K.A. Prather, T. Reiner, H. Rodhe, H.A. Scheeren, D. Sikka, and J. Williams; The Indian Ocean Experiment: Widespread Air Pollution from South and Southeast Asia, Science, 291 (2001) 1031-1036. Magneron, I., R. Thevenet, A. Mellouki, G. Le Bras, G.K. Moortgat, and K. Wirtz; A Study of the Photolysis and OH-initiated Oxidation of Acrolein and trans-Crotonaldehyde, Journal of Physical Chemistry A, 106 (2002) 2526-2537. Middleton, J.T., J.B. Kendrick, and H.W. Schwalm; Injury to Herbaceous Plants by Smog or Air Pollution, USDA Plant Dis.Rep., 34 (1950) 245-252.

Remote Sensing of Glyoxal by Differential Optical Absorption Spectroscopy

141

Molina, M.J. and L.T. Molina (Eds.) (2002), Air Quality in the Mexico Megacity: An Integrated Assessment, Kluwer Academic Publishers, Dordrecht, The Netherlands. Moortgat, G.K., D. Grossmann, A. Boddenberg, G. Dallmann, A.P. Ligon, W.V. Turner, S. Gab, F. Slemr, W. Wieprecht, K. Acker, M. Kibler, S. Schlomski, and K. Bachmann; Hydrogen peroxide, organic peroxides and higher carbonyl compounds determined during the BERLIOZ campaign, Journal of Atmospheric Chemistry, 42 (2002) 443-463. Paulson, S.E. and J.H. Seinfeld; Development and Evaluation of A Photooxidation Mechanism for Isoprene, Journal of Geophysical Research-Atmospheres, 97 (1992) 20703-20715. Platt, U. (1994), Differential Optical Absorption Spectroscopy, in Monitoring by Spectroscopic Techniques, edited by Sigrist, M.W., chapter 2, 27-84, Wiley & Sons, New York. Plum, C.N., E. Sanhueza, R. Atkinson, W.P.L. Carter, and J.N. Pitts; Oh Radical Rate Constants and Photolysis Rates of Alpha-Dicarbonyls, Environmental Science & Technology, 17 (1983) 479-484. RADICAL; Evaluation of Radical Sources in Atmospheric Chemistry through Chamber and Laboratory studies, ENV4-CT97-0419 (2001) Ramanathan, V. and P.J. Crutzen; New directions: Atmospheric brown "Clouds", Atmospheric Environment, 37 (2003) 4033-4035. Saunders, S.M., M.E. Jenkin, R.G. Derwent, and M.J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3 (Part A): tropospheric degradation of non-aromatic volatile organic compounds, Atmospheric Chemistry and Physics, 3 (2003) 161-180. Smith, D.F., C.D. Mciver, and T.E. Kleindienst; Primary product distribution from the reaction of hydroxyl radicals with toluene at ppb NOX mixing ratios, Journal of Atmospheric Chemistry, 30 (1998) 209-228. Volkamer, R. (2001), A DOAS Study on the Oxidation Mechanism of Aromatic Hydrocarbons under Simulated Atmospheric Conditions, Dissertation, University of Heidelberg, Germany http://archiv.ub.uniheidelberg.de/volltextserver/volltexte/2002/1939/ Volkamer, R., B. Klotz, I. Barnes, T. Imamura, K. Wirtz, N. Washida, K.H. Becker, and U. Platt; OH-initiated oxidation of benzene - Part I. Phenol formation under atmospheric conditions, Physical Chemistry Chemical Physics, 4 (2002) 1598-1610. Volkamer, R., L.T. Molina, M.J. Molina, T. Shirley, and W.H. Brune; DOAS measurement of glyoxal as an indicator for fast VOC chemistry in urban air, Geophysical Research Letters, 32 (2005a) L08806, doi: 10.1029/2005GL022616. Volkamer, R., U. Platt, and K. Wirtz; Primary and Secondary Glyoxal Formation from Aromatics: Experimental Evidence for the Bicycloalkyl-Radical Pathway from Benzene, Toluene, and p-Xylene, Journal of Physical Chemistry A, 105 (2001) 7865-7874. Volkamer, R., P. Spietz, J.P. Burrows, and U. Platt; High-resolution absorption cross-section of Glyoxal in the UV/vis and IR spectral ranges, Journal of Photochemistry and Photobiology A: Chemistry, 172 (2005b) 3542, doi: 10.1016/j.jphotochem.2004.11.011.

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies Claire Bloss1, Michael E. Jenkin2, William J. Bloss1, Andrew R. Rickard1, and Michael J. Pilling1 2

1 School of Chemistry, University of Leeds, Leeds LS2 9JT, UK Imperial College London, Silwood Park, Ascot, Berkshire SL5 7PY, UK

Key Words: Aromatic, Oxidation, Mechanism, Troposphere, Ozone, Chamber

Abstract The Master Chemical Mechanism (http://mcm.leeds.ac.uk/MCM) has been updated to version MCMv3.1 in order to incorporate recent kinetic and mechanistic improvements in the understanding of aromatic photo-oxidation. Carefully designed experiments have been carried out in the European Photoreactor (EUPHORE) to investigate key subsets of the toluene system. These results have been used to evaluate the toluene mechanism in MCMv3.1 (and by extension the other aromatic mechanisms in MCMv3.1) and where appropriate to refine the degradation schemes. MCMv3 and MCMv3,1 have also been evaluated using a high quality EUPHORE dataset on the photo-oxidation of benzene, toluene, p-xylene and 1,3,5-trimenthylbenzene. Significant deficiencies have been identified in the mechanisms, in particular: (1) an over-estimation of the ozone concentration, (2) an under estimation of the NO oxidation rate and (3) an under-estimation of OH. The use of MCMv3.1 improves the model-measurement agreement in some areas but significant discrepancies remain. Ideas for additional modifications to the mechanisms and for future experiments to further our knowledge of the details of aromatic photo-oxidation are discussed. Introduction It has been estimated that aromatic compounds, whose major sources are from petrol vehicle exhaust and solvent usage, contribute ~10% to global anthropogenic non-methane hydrocarbon emissions. In regions where there are relatively strong traffic emissions, this percentage can rise substantially (c.f. 21% in the UK (PORG (1997)). The contribution of aromatics to regional photochemical ozone formation is of even greater importance owing to their high reactivity (and hence high photochemical ozone creation potentials (POCPs)) and they can form significant amounts of secondary organic aerosol. For example, in the UK aromatics contribute c.a. 30% to regional ozone production (Derwent et al. (1996)). Therefore, it is important to understand the detailed photochemical degradation mechanisms of aromatic compounds, under a variety of atmospheric conditions, for use in air quality models. The degradation schemes of four aromatic hydrocarbons; benzene, toluene, p-xylene and 1,3,5-trimethylbenzene, have been updated on the basis of new kinetic and mechanistic data from current literature and conference proceedings and are available as part of the latest version of the Master Chemical Mechanism (MCMv3.1) via the MCM website (http://mcm.leeds.ac.uk/MCM). The performance of these schemes concerning ozone formation from tropospheric aromatic oxidation has been evaluated using detailed environmental chamber datasets from the two EU EXACT measurement campaigns at EUPHORE (EXACT I – September 2001 and EXACT II – July 2002 (Pilling et al., 2003)). 143 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 143–154. © 2006 Springer. Printed in the Netherlands.

C. Bloss et al.

144

This comprehensive databank includes photochemical smog chamber studies on the four classes of mono-aromatic compounds above (carried out under suitable atmospheric conditions) as well as a series of experiments looking into specific key areas of aromatic photo-oxidation, focussing on subsets of the toluene system. Results from this system can be transferred to other aromatics as the mechanisms are expected to be similar. Where appropriate, results from this database have been used to refine the mechanisms. Experimental The aromatic validation experiments were performed in chamber A during two campaigns at EUPHORE. A range of techniques was used to measure the concentrations of important species in each system. To ensure high data quality the parent aromatic and ozone were measured with two independent methods, FTIR/GC and UV/FTIR, respectively. NO2 concentrations were measured using DOAS and NO was continuously monitored by a chemiluminescence analyser. LIF was used to measure the concentration of OH radicals during both campaigns, and also measured HO2 during EXACT II (Bloss et al., 2003a). FTIR was used to monitor formaldehyde and nitric acid, GC-ECD was used for PAN and methylglyoxal measurements, DOAS measured glyoxal in addition to NO2, and HPLC detection was used for ring-retained products such as cresols, benzaldehyde and benzoquinones. A comprehensive GC technique from the University of Leeds was used to monitor products such as -dicarbonyls and furanones expected in the ring-opening routes of aromatic oxidation (Hamilton et al., 2003). j-(NO2) was measured using a filter radiometer and this measurement was used to scale all other photolysis rates calculated in the models. Results Development of a detailed chemical mechanism (MCMv3.1) for the atmospheric oxidation of aromatic hydrocarbons The MCM is a near-explicit chemical mechanism developed to describe, in detail, the tropospheric degradation of various important emitted VOCs and is available through the MCM website (http://mcm.leeds.ac.uk/MCM). The first versions of the mechanism used simplified provisional schemes for the poorly known chemistry of aromatic VOC, but in recent years there has been a number of laboratory studies on the atmospheric oxidation of aromatic hydrocarbons and the understanding of the detailed chemistry has improved (e.g. Calvert et al. (2002)). This information has been used to define a rigorous and detailed protocol for the construction of aromatic degradation mechanisms, and these are included in MCM version 3 (Jenkin et al., 2003). However, significant uncertainties remain and development work on these schemes is ongoing. The schemes for benzene, toluene, p-xylene and 1,3,5-trimethylbenzene have been updated following the evaluation experiments performed during EXACT and other data which has become available since the launch of MCMv3. These updated aromatic schemes are now available (via the website) in MCMv3.1. The oxidation pathways of toluene as implemented in the original mechanism (MCMv3) and the updated mechanism (MCMv3.1) are shown in Figure 1. In general, the key areas in which the aromatic mechanisms have been changed are listed below (a more detailed discussion on the updated aromatic chemistry in MCMv3.1 is available in Bloss et al., 2005b):

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies

145

x Benzaldehyde yield is slightly lower in the toluene system in MCMv3.1 than MCMv3 in order to better represent the more recent data at low NOx levels (Calvert et al., 2002). x Photolysis rates and subsequent product branching ratios of unsaturated -dicarbonyls (an important source of radicals) have been adjusted according to Thuener et al. (2003) for butenedial and 4-oxo-2-pentenal, and Graedler and Barnes (1997) for 3-hexene-2,5-dione. The photolysis rates of the epoxydicarbonylene products have also been increased as they were originally estimated by analogy with the unsaturated -dicarbonyls. x The breakdown of the carbon skeleton for 2(5H)-furanone (also a photolysis product of butenedial) has been updated and -angelica lactone has been replaced by -angelica lactone to reduce secondary glyoxal formation. These changes were prompted by findings of Volkamer et al. (2001) that secondary glyoxal formation in the toluene system is negligible. x New phenol-type chemistry has been implemented reflecting the lower yield for ringopening in this channel reported by Olariu et al. (2002), and the need for reduced ozone formation found in a cresol photosmog experiment at EUPHORE (e.g. see Figure 5). x Branching ratios for the different oxidation routes of aromatics have been adjusted to reflect the reported yields of glyoxal and of phenol-type compounds at NOx levels appropriate to the atmosphere (Volkamer et al., 2001; Volkamer et al., 2002).

Figure 1.

Schematic representations of toluene oxidation mechanisms. Upper panel MCMv3, lower panel MCMv3.1.

A sensitivity analysis carried out on the MCMv3.1 aromatic chemistry identified the Jdicarbonyls intermediates (co-products of glyoxals in the peroxy-bicyclic ring opening step in MCMv3.1, see Figure 1) as important species affecting OH and O3 concentrations in these systems (Bloss et al., 2005b). 3A comparison of modelled and measured concentrations from a photosmog experiment on butenedial is shown in Figure 2. This experiment and similar

146

C. Bloss et al.

studies on 4-oxo-2-pentenal indicated a number of shortcomings in the mechanisms. NO2 concentration-time profiles are poorly simulated for all experiments, and the differences between experiments under different initial conditions are not well represented. It is clear that the NOxy budget is not well understood. The removal of butenedial is well simulated by MCMv3.1 but the simulated OH and HO2 radical concentrations are significantly lower than the observed values and it seems that a large radical source is missing from the mechanism. This important issue so far remains unresolved. Figure 2 shows a model/measurement comparison from a butenedial photolysis experiment in the absence of NOx. The loss of butenedial is well predicted by MCMv3.1. However, the HO2 concentration is over-estimated by MCMv3.1 by almost an order of magnitude during the early part of the experiment. The time-dependent behaviour is also not well reproduced by the simulation as in the experiment an initial fast increase in concentration is followed by a slower linear increase until the chamber closes, while the simulation shows a fast rise followed by a fall in the HO2 concentration even while the photolysis continues. The photolysis mechanism for butenedial in the absence of NOx as implemented in MCMv3.1 is shown schematically in Figure 4. This indicates that two HO2 radicals should be formed for each molecule of maleic anhydride and glyoxal produced, and while both these product concentrations are over-estimated this is not sufficient to account for the large over-prediction of HO2.

Figure 2. Butenedial photosmog experiment: Figure 3. Butenedial photolysis experiment: comparison of modelled and measured comparison of modelled and measconcentrations of parent compound, ured concentrations of butenedial, HO2, and selected products. O3 , NOx and HOx .

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies

Figure 4.

147

Schematic representation of MCMv3.1 butenedial photolysis mechanism in the absence of NOx .

Thuener et al. (2003) propose an alternative butenedial photolysis mechanism based on product experiments. However, this mechanism also requires two HO2 radicals per maleic anhydride molecule formed. One must conclude that a different HO2 formation pathway is operating and/or a significant loss process for HO2 is not included in the mechanism. Concerning the molecular products of butenedial photolysis (Figure 3), the yield of 2(5H)-furanone is well predicted by the simulation, glyoxal and maleic anhydride are overpredicted while the CO yield in the simulation is much lower than observed experimentally. In MCMv3.1 glyoxal and CO are formed as co-products (Figure 4), but this is not consistent with the different yields of these products observed experimentally. Thuener et al. (2003) include a different source of CO in their proposed mechanism, i.e. direct formation from photolysis with a yield of 20%. A possible co-product for direct CO production is acrolein formed by an H-shift and C-C cleavage. The acrolein concentration was below the detection limit of the measurement technique, and its maximum yield was estimated to be ~ 10%. No other direct photolysis products were observed and it was not possible to positively determine the mechanism and co-products for CO formation. In addition to the products for which results are shown in Figure 3, 4-Oxo-but-2-enoic acid and 4-Oxo-but-2-eneperoxoic acid are predicted by MCMv3.1 to be major products in the NOx free conditions of this experiment but were not observed. An unattributed carbonyl absorption was found in the FTIR spectra (Thuener et al., 2003) but the absorption frequency of this unknown does not correspond to that expected for a conjugated carboxylic or peroxoic acid. Although some improvements to our understanding of the photo-oxidation of Jdicarbonyls have been made during the EXACT project the agreement between MCM simulations and chamber experiments, both with and without NOx, remains poor. Particular problems arise concerning the radical yields and seemingly contradictory behaviour is observed. With NOx present, the simulations under-estimate both OH and HO2 while in the absence of NOx the HO2 concentration is over-estimated. In the photosmog systems the NOy budget is not well understood and the simulations are unable to reproduce the different

148

C. Bloss et al.

behaviours seen experimentally at different VOC/NOx ratios and with different J-dicarbonyls. It is therefore difficult to identify mechanistic changes which can simultaneously resolve the discrepancies seen in the two systems (i.e. photosmog and photolysis), based on current general understanding of processes involved in free radical driven VOC oxidation. Nevertheless, overcoming these shortcomings in an area of the mechanism identified as being of key importance in determining ozone and OH concentrations is an important issue for improving our understanding of aromatic chemistry. Another problem encountered in the aromatic photosmog experiments was the difficulty experienced in detecting appropriate amounts of J-dicarbonyls under atmospheric conditions, even using a comprehensive 2D-GC technique (Hamilton et al., 2003). Therefore, improved detection techniques need to be investigated in order to quantify the amounts of J-dicarbonyls formed in these systems. Nevertheless, the EUPHORE chamber dataset on unsaturated Jdicarbonyls provides an important resource for testing future mechanism developments and the work presented here highlights issues that require further investigation. Evaluation of detailed aromatic mechanisms against environmental chamber data As already mentioned, a large, high quality smog chamber dataset on the photooxidation of benzene, toluene, p-xylene and 1,3,5-trimethylbenzene has been assembled from experiments carried out as part of the two EUPHORE EXACT campaigns. The experiments were designed to test sensitive features of detailed aromatic mechanisms, and the dataset has been used to evaluate the performance of the MCMv3 and the updated MCMv3.1. Detailed descriptions of the experimental design and procedure are given in Bloss et al. (2003a) and (2005a). Table 1 summarises the model to measured performance for each experiment in the EXACT photo-smog database. The investigated aromatics form a series of compounds with increasing methyl substitution, and for each compound a range of NOx/VOC ratios are covered by the experiments. The ethene results are included in order to show the level of agreement seen for a relatively well understood system. Figures 5 and 6 show examples of aromatic model-measurement comparisons for benzene and toluene respectively. The O3 peak in the benzene system is greatly reduced in MCMv3.1 compared to MCMv3, and is in good agreement with the measurements, as shown in Figure 5. This substantial improvement is mainly a result of an increased phenol channel yield implemented in MCMv3.1 and an increased ring-retaining product yield in that channel. However, this increase in ring-retaining yield and consequent decrease in radical production leads to a reduction in OH production and oxidative capacity of the system (which was better simulated by MCMv3) as shown in the missing OH calculations in Table 1 and in the decay of benzene in Figure 5. This is indicative of a general problem with the mechanisms; overprediction of the ozone concentration but under-prediction of the reactivity of the system. For the toluene system, model-measurement agreement in terms of O 3 concentration and NO oxidation is improved in MCMv3.1 but the maximum O3 is still over estimated as shown in Figure 6. This mechanism has an increased branching ratio for ring opening in toluene oxidation, and a slightly decreased cresol yield (see Figure 1). These changes, along with the increased photolysis rates of the unsaturated dicarbonyl compounds, tend to increase radical production, particularly early in the experiment, and hence increase O3 generation. However, changes to the substituted phenol (cresol) chemistry, analogous to those discussed above for the benzene system, decrease O3 formation and reduce radical production in the middle of the experiment. The changes balance out in such a way that some decrease in O3

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies

149

formation is achieved while increasing the radical production in the early stages of the experiment. However, the amount of “missing” OH needed for the model to reproduce the observed decay of the primary aromatic was calculated over the whole course of the experiment, and in general is higher for the MCMv3.1 simulations indicating an overall reduction in oxidative capacity for this mechanism consistent with the reduced ozone formation potential (see Table 1). Table 1. Measures of model performance concerning the ozone formation and oxidation capacity for MCMv3 and MCMv3.1. For O3 the peak values are taken irrespective of the time they occur, NO oxidation rate is a measure for the timing of the simulation, and the numbers in bold for these measures are the percentage difference between simulated and observed value, relative to the observed value. For missing OH, base value is the amount of new OH produced by the base mechanism over the course of the experiment, “missing” is the amount of extra OH needed to achieve agreement between modelled and observed aromatic decay, and the number in bold is the missing OH as a percentage of the total OH, (base + missing).

In general the updated mechanism, MCMv3.1, shows improved ability to simulate some of the observations from the EXACT EUPHORE datasets and represents our current understanding of aromatic degradation. However, significant discrepancies remain concerning, in particular, ozone formation potential and oxidative capacity of aromatic hydrocarbon systems: x The peak ozone concentration in the benzene experiments is well simulated but for the substituted aromatics the simulations generally over-estimate the peak ozone concentration. x The OH radical production in the mechanisms is too low to account for the OH concentration inferred from the rate of loss of the parent aromatic. The simulations underestimate the oxidative capacity of the aromatic systems. x For the majority of the experiments the simulations under-predict the NO oxidation rate, which is used as a measure of the timing of the simulations. This parameter is closely

150

C. Bloss et al.

related to the processes responsible for the production of ozone and the reactivity of the system.

Figure 5. Model-measurement comparison for Figure 6. Model-measurement comparison for toluene, O3, NO2 and NO in a benzene, O 3 , NO2 and NO in a EUPHORE toluene photosmog EUPHORE benzene photosmog experiment. experiment.

The simultaneous over prediction of O3 concentrations and under prediction of the NO oxidation rate and oxidising capacity of the aromatic systems poses a problem for mechanism development; a reduction in ozone concentration can be achieved by a reduction in peroxy radical concentration, which limits the NO to NO2 conversion, but this will also lead to a reduction in OH production and in the oxidative capacity of the system. A number of possible mechanistic fixes have been investigated and a number of strategies for resolving some of these issues have been proposed (Bloss et al., 2005b): x An increase in OH yield without additional NO to NO2 conversion would improve the simulated concentration-time profiles. A potential OH regeneration channel has been identified, but there is no firm evidence for its occurrence. x Conversion of NO3 to HO2 by a dummy reaction in the model decreases O3 yields but does not significantly affect the aromatic decay rate. x A modelled conversion of NO2 to HONO on the secondary organic aerosol has been shown to improve model-measurement agreement for O3, NOx and toluene decay. For example, Figure 7 shows the effect on the simulations of replacing the usual peroxy radical chemistry by a direct decomposition to ring-opened products and OH. This could occur following a rearrangement of the peroxy radical by a H-shift, as indicated in Figure 8.

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies

151

Figure 7. Comparison between experimental Figure 8. Schematic of MCMv3.1 peroxide bicyclic route for toluene oxidation and simulated concentration timeand postulated OH regeneration profiles for MCM3.1 and mechanism pathway. including a speculative OH regeneration mechanism in the peroxidebicyclic route for an EXACT Toluene experiment

It is clear to see that this alteration in the mechanism improves the modelmeasurement agreement. The peak ozone is decreased by 12% compared to MCMv3.1, and the measured peak concentration is over-predicted by 19% by the modified mechanism, 36% by MCMv3.1. At the same time the modelled loss of toluene is faster than in MCMv3.1 as a result of higher OH concentrations, giving a better representation of the observations but still an under-estimate of the reactivity of the system. In the aromatic chamber experiments aerosol particles are formed and chemistry may take place on the surface of these particles. For example, NO2 could partition to the condensed phase and be reduced to HONO by the surface bound species on the secondary organic aerosol formed in the aromatic chamber experiments. Such heterogeneous processes would provide a sink for NOx and a source of radicals (via the photolysis of HONO). A comparison between experimental and simulated concentration time-profiles for MCM3.1 and mechanisms including a conversion of NO2 to HONO on the measured aerosol surface for the toluene experiment shown in Figure 6 is shown in Figure 9. The rate of NO2 reaction on the aerosol was calculated as a function of time, assuming a constant reaction probability of 0.025. The total surface area of the aerosol is also shown and a sharp onset of particle formation was observed just after 11:00. At this point the simulations show a decrease in NO2 concentration and increase in HONO concentration. The OH concentration also increases rapidly, as evidenced by the increase in toluene loss rate. The peak ozone concentration is decreased and under-estimates the measured value. However, this underestimation is a result of the low reactivity in the early part of the simulation before aerosol formation takes place. If an OH source is introduced to reproduce the toluene loss in the first hour of the simulation the NOx and O3 concentrations agree well with the measurements, as shown in Figure 9. However, while the simulations predict a large peak in HONO concentrations no evidence for HONO formation was seen in the FTIR spectra suggesting that the concentration was below approximately 6 ppbv.

152

C. Bloss et al.

Figure 9. Comparison between experimental and simulated concentration time-profiles for MCM3.1, mechanism with a dummy NO3 conversion to HO2 included and mechanisms with artificial OH source added in order to reproduce toluene loss for an EXACT Toluene experiment Summary and Conclusions The detailed oxidation mechanisms of aromatic hydrocarbons in the MCM have been revised and updated based on the latest available experimental data and is available via the MCM website (http://mcm.leeds.ac.uk/MCM) as MCMv3.1. A series of chamber experiments were carried out to investigate the details of key areas of aromatic oxidation mechanisms, and these were primarily focused on toluene oxidation (Bloss et al., 2005b). Comparison of modelled and measured concentrations from photosmog experiments on J-dicarbonyls indicated a number of shortcomings in the mechanisms. NO2 concentrationtime profiles are poorly simulated for all experiments, and the differences between experiments under different initial conditions are not well represented. It is clear that the NOxy budget is not well understood. The simulated HOx concentrations are significantly lower than the observed values and it seems that a large radical source is missing from the mechanism. However, very low HO2 radical concentrations were measured in J-dicarbonyl photolysis experiments in the absence of NOx, these measured concentrations were much lower than HO2 concentrations simulated by MCMv3.1. These important issues remain unresolved. More experiments are needed to determine the yields of reactive oxygenated intermediates formed in aromatic degradation. Ideally these would be carried out with high time resolution as an aid to distinguishing primary and secondary oxidation products. Sensitive, on-line analytical techniques, such as proton-transfer-reaction mass spectrometry, should be employed to detect and quantify such intermediates in chamber experiments carried out under NOx conditions representative of atmospheric levels. Experiments have been carried out in EUPHORE to study the oxidation of benzene, toluene, p-xylene and 1,3,5-trimethylbenzene under conditions appropriate to the atmosphere, and the large, high quality dataset obtained has been used to evaluate the performance of the

Aromatic Hydrocarbon Oxidation: The Contribution of Chamber Oxidation Studies

153

MCMv3 and MCMv3.1 aromatic mechanisms (Bloss et al., 2005a). The updated mechanism, MCM v3.1, shows improved ability to simulate some of the observations from the EUPHORE dataset but significant discrepancies remain, concerning, in particular, ozone formation potential and oxidative capacity of aromatic hydrocarbon systems. The combination of over-prediction of ozone concentration and under-prediction of reactivity poses a problem for mechanism development; a reduction in ozone concentration can be achieved by a reduction in peroxy radical concentration, which limits the NO to NO2 conversion, but this will also lead to a reduction in NO oxidation rate and OH production. A number of possible mechanistic fixes have been investigated (Bloss et al., 2005b) as listed below: x A regeneration mechanism for OH formation without additional NO to NO2 conversion would improve the simulated concentration-time profiles. Flash photolysis experiments with time-resolved OH detection by laser induced fluorescence could be carried out to determine radical yields in the initial stages of aromatic oxidation. The hydroxycyclohexadienyl radical, which in the atmosphere is formed by OH addition to the aromatic ring, could be generated by photolysis of a suitable precursor compound. If the postulated OH regeneration mechanism illustrated in Figure 8 were correct, subsequent reactions with O2 and intramolecular rearrangements would lead to production of OH in the absence of NO. This experiment would require synthesis of a suitable photolytic precursor for the hydroxycyclohexadienyl radical, and careful consideration of the appropriate initial precursor and O2 concentrations. x Conversion of NO3 to HO2 by a dummy reaction in the model decreases O3 yields but does not significantly affect the aromatic decay rate. This indicates that some NO3 chemistry may be missing from the mechanism and could help to explain some of the model-measurement discrepancies. However, any such chemical modifications will not solve the problem of the under-estimation of the oxidative capacity in the system. x A modelled conversion of NO2 to HONO on the secondary organic aerosol has been shown to improve model-measurement agreement for O3, NOx and toluene decay. However, the relatively high HONO concentrations generated have not been observed experimentally and the reactive uptake coefficient required is much higher than the upper limits for such processes suggested by the work of Broske et al. (2003). Further investigation of radical and NOxy budgets under conditions of significant organic aerosol concentration is required to determine the effect of heterogeneous reactions in these aromatic systems. x A sensitivity analysis is planned, to investigate the effect of the uncertainty in the rate of radical generation on ozone formation in the standard photochemical trajectory model (Derwent et al. (1996)). In an environment with many VOC species and radical precursors, this uncertainty may be less significant. The development work discussed here on aromatic degradation schemes has been extended by analogy to update the degradation schemes of twelve other mono-aromatics in the MCMv3.1 with saturated alkyl side chains.

154

C. Bloss et al.

Acknowledgements The authors would like to thank Klaus Wirtz and Montse Martin-Reviejo for their management skills and scientific contributions during the EXACT EUPHORE campaigns and the rest of the EXACT consortium for all their substantial contributions. This work was supported by the European Commission within the EXACT project, Contract No. EVK2-CT1999-00053. References Bloss, C., V. Wagner, A.Bonzanini, M.E. Jenkin, K. Wirtz, M. Martin-Reviejo and M.J. Pilling; Evaluation of the master chemical mechanism (MCM) for aromatic hydrocarbons., EUPHORE Annual Report 2001. (2003a) 34-40. Bloss, C., V. Wagner, A. Bonzanini, M.E. Jenkin, K. Wirtz, M. Martin-Reviejo and M.J. Pilling; Evaluation of detailed aromatic mechanisms (MCMv3 and MCMv3.1) against environmental chamber data., Atmos. Chem. Phys., 5, (2005a) 623-639. Bloss, C., V. Wagner, M.E. Jenkin, R. Volkamer, W.J. Bloss, J.D. Lee, D.E. Heard, K. Wirtz, M. MartinReviejo, G. Rea, J.C. Wenger and M.J. Pilling; Development of a detailed chemical mechanism (MCMv3.1) for the atmospheric oxidation of aromatic hydrocarbons. Atmos Chem. Phys., 5, (2005b) 641-664. Bloss, W.J., J.D.Lee, C. Bloss, K. Wirtz, M. Martin-Reviejo, M. Siese, D.E. Heard and M.J. Pilling; Validation of the calibration of a laser-induced fluorescence instrument for the measurement of OH radicals in the atmosphere. Atmos. Chem. Phys., 4, (2003b) 571-583. Broske, R., J. Kleffmann and P. Wiesen; Heterogeneous conversion of NO2 on secondary organic aerosol surfaces: A possible source of nitrous acid (HONO) in the atmosphere? Atmos. Chem. Phys. 3, (2003) 469474. Calvert, J.G., R. Atkinson, K.H Becker, R.M. Kamens, J.H. Seinfeld, T.J. Wallington and G. Yarwood; The Mechanisms of Atmospheric Oxidation of Aromatic Hydrocarbons, Oxford University Press (2002). Derwent, R.G., M.E. Jenkin and S.M. Saunders; Photochemical ozone creation potentials for a large number of reactive hydrocarbons under European conditions. Atmos. Environ. 30(2), (1996) 181-199. Graedler, F. and I. Barnes; Photolysis of Z-/E-3-Hexenedione. EUPHORE Annual Report 1997, (1997) 148. Hamilton, J.F., A.C. Lewis, C. Bloss, V. Wagner, A.P. Henderson, B.T. Golding, K. Wirtz, K, M. MartinReviejo and M.J. Pilling; Measurements of photo-oxidation products from the reaction of a series of alkylbenzenes with hydroxyl radicals during EXACT using comprehensive gas chromatography. Atmos. Chem. Phys., 3, (2003) 1999-2014. Jenkin, M.E., S.M. Saunders, V. Wagner and M.J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3 (Part B): tropospheric degradation of aromatic volatile organic compounds. Atmos. Chem. Phys. 3, (2003) 181-193. Olariu, R.I., B. Klotz, I. Barnes, K.H. Becker and R. Mocanu; FT-IR study of the ring-retaining products from the reaction of OH radicals with phenol, o-, m-, and p-cresol. Atmos. Environ. 36(22), (2002) 3685-3697. Pilling, M.J. et al.; Effects of the oXidation of Aromatic Compounds in the Troposphere (EXACT, EVK2-CT1999-00053), Final Report., http://www.chem.leeds.ac.uk/exact, (2003). PORG 1997; "Ozone in the United Kingdom", Report prepared by the UK Photochemical Oxidants Review Group. London, HMSO Publications, (1997). Thuener, L.P., G. Rea, and J.C. Wenger; Photolysis of butenedial and 4-oxo-2-pentenal., EUPHORE Annual Report 2001, (2003) 41-46. Volkamer, R., U. Platt, and K. Wirtz; Primary and secondary glyoxal formation from aromatics: Experimental evidence for the bicycloalkyl-radical pathway from benzene, toluene, and p-xylene. J. Phys. Chem. A. 105(33), (2001) 7865-7874. Volkamer, R., B. Klotz, I. Barnes, T. Imamura, K. Wirtz, N. Washida, K.H. Becker and U. Platt; OH-initiated oxidation of benzene - Part I. Phenol formation under atmospheric conditions. Phys. Chem. Chem. Phys. 4(9), (2002) 1598-1610.

FT-IR Kinetic Study on the Gas-Phase Reactions of the OH Radical with a Series of Nitroaromatic Compounds Iustinian Bejan1,2, Ian Barnes1, Romeo Olariu2, Karl Heinz Becker1, and Raluca Mocanu2 1

Bergische Universität Wuppertal, FB-C - Physikalische Chemie, Gaußstraße 20, D-42097 Wuppertal, Germany “Alexandru Ioan Cuza” University of Iasi, Faculty of Chemistry, Department of Analytical Chemistry, Carol I Boulevard 11, 700506 Iasi, Romania

2

Key Words: Rate constant, Gas-Phase, OH Radical, Photolysis, Nitroaromatic, OH formation

Introduction Large quantities of harmful VOCs are emitted into the troposphere from anthropogenic sources (Calvert et al., 2002). Aromatic hydrocarbons are an important class of VOCs present in the atmosphere, which contribute significantly to the chemistry of urban air (Atkinson, 2000; Atkinson and Arey, 2003). Estimations of the global emissions of aromatic hydrocarbon suggest that they comprise between 17-25% of the total anthropogenic NMVOC emissions (Calvert et al., 2002). The degradation of aromatic hydrocarbons is mainly initiated during the day by reaction with the hydroxyl radical (OH). Based on the presently accepted but inadequate mechanism for the atmospheric degradation of aromatic hydrocarbons, it has been calculated that this class of VOC could account for up to 30% of the photooxidant formation in urban areas (Derwent et al., 1996, 1998). Besides the photooxidant formation, this class of hydrocarbon is also assumed to make a significant contribution to secondary organic aerosol (SOA) formation in urban areas (Odum et al., 1996; Forstner et al., 1997; Hurley et al., 2001). This ranks aromatic hydrocarbon as one of the most important classes of hydrocarbons emitted into the urban atmosphere. Formulation of a realistic photo-oxidation mechanism for aromatic hydrocarbons represents one of the most important and challenging problems remaining to be solved in chemical models of tropospheric photooxidant formation. Nitroaromatics such as nitrophenols and nitrocresols have been measured in significant quantities in the atmosphere (Atkinson et al., 1992; Grosjean, 1991), mainly in relation to their potential health effects. Nojima et al. (1975) were the first to detect nitrophenols in the environment and reported the presence of several nitrophenols in the rain water. Nitroaromatics have gained more in atmospheric importance since their phytotoxic properties were discovered (Kawai et al., 1987). Nitroaromatics have been identified in air (Belloli et al., 1999; Tremp et al., 1993), clouds (Lüttke and Levsen, 1997), water (Geissller and Scholer, 1994), soil (Voznakova et al., 1996), fog (Herterich, 1991) and snow (Kawamura and Kaplan, 1986). Nitrated hydroxyaromatics may enter into the atmosphere from both primary and secondary sources. The formation of nitrophenols and nitrocresols in the combustion processes of motor vehicles has been reported by Tremp et al. (1993). Others primary sources may be combustion of coal, wood, manufacture of phenol-formaldehyde resins, pharmaceuticals disinfectants, dyes and explosives (Harrison et al., 2005). Studies in our and other laboratories have shown that an additional important source of these compounds in the atmosphere could be the gas-phase OH-radical initiated photooxidation of aromatic hydrocarbons such as benzene, toluene, phenol, cresols and dihydroxybenzenes in the presence of NOx during the daytime as well as the reaction of NO3 radicals with these aromatics during the night time (Atkinson et al., 1992; Olariu et al., 2002). Once released or 155 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 155–162. © 2006 Springer. Printed in the Netherlands.

I. Bejan et al.

156

produced in the troposphere, these compounds will undergo further oxidation and are also expected to partition between the gas and aerosol phases. Presently very little is known about the atmospheric behaviour of nitrophenols and nitrocresols. The aim of the present study is to increase our knowledge with respect to the fate of this compound class under simulated tropospheric conditions in the laboratory. Recent studies in our laboratory on the photolysis of nitrophenol and nitrocresols with actinic lamps (Philips TL 05/40 W: 320 < Ȝ < 480, Ȝmax = 360 nm) have indicated that these compounds could be important secondary organic aerosol sources (Bejan et al., 2003). In order to assess the importance of this aerosol formation pathway for the atmosphere, kinetic information concerning other atmospheric loss processes, e.g. reactions with OH and NO3 radicals, is required. Experimental The kinetic investigations were carried out in a 1080 L cylindrical quartz-glass photoreactor in synthetic air (296 ± 3 K and 760 ± 10 Torr). As the description given here is limited to the instrumentation actually used in this study, the reader is referred to the literature for a more detailed description of this reaction chamber (Barnes et al., 1994). A White type multiple reflection mirror system, operated at a total optical path length of (484.7 ± 0.8) m, coupled to a Nicolet Nexus FTIR spectrometer equipped with a liquid nitrogen cooled mercury-cadmium-tellurium (MCT) detector, was used for reactant and reference monitoring. IR spectra were recorded at a spectral resolution of 1 cm-1. Between 8 and 12 spectra were recorded in each kinetic experiment, every spectrum comprised of 128 co-added interferograms. Before the kinetic experiment a series of five experiments were performed for every nitrocresol compound in order to obtain the photolysis and wall loss rates. These latter two constants were used in corrections of the OH kinetic rate data. Formation of OH radicals was observed during the photolysis of nitrocresols. Using isoprene as scavenger for OH radicals (kOH = 1.0 × 10-10 cm3 s-1, Atkinson and Arey, 2003) it was possible to calculate the concentration of OH radicals produced by the photolysis and thus calculate a correct rate constant for the photolysis frequency. The photolysis of CH3ONO with 32 superactinic fluorescent lamps (Philips TL05/40 W; 320 nm < O < 480 nm, with Omax=360 nm) in the presence of NO was used as the OH radical source: CH3ONO

+

hQ

CH3O +

CH3O

+

O2

HCHO +

HO2

+

NO

OH

+

NO

HO2 NO2

Gaseous compounds (CH3ONO, NO, ethene, n-butane and isoprene) were injected into the reactor using calibrated gas-tight syringes with the reactor under reduced pressure; solid compounds (nitrocresols) were introduced into the chamber in an air flow through a special heated inlet system. The initial reactant concentrations (in molecules cm-3 units) were as follows: C5H8 2× 1014, C4H10 (6-8) × 1013, nitrocresols (4.8 - 9.2) × 1013, CH3ONO 5 × 1013, C2H4 (6 – 8) × 1013 and NO 3 × 1014.

157

FT-IR Kinetic Study on the Gas-Phase Reactions of the OH Radical

6-Methyl-2-nitrophenol was prepared by addition of sodium nitrite to o-toluidine dissolved in a concentrated solution of sulphuric acid (Winzor, 1935). The rate constants were determined using the relative kinetic technique. Ethene was used as the reference compound. The OH radicals generated in the photolysis systems will react with both the nitrocresols and the reference hydrocarbon: reactant + OH

o products

(k1)

(1)

reference + OH

o products

(k2)

(2)

Under the experimental conditions, the aromatic compounds are also lost by absorption on the surface of the reactor and by photolysis. Control experiments have been performed in order to measure these loss rates: reactant (+wall loss, hQ)

o products

(k3)

(3)

No wall deposition or photolysis was observed for the reference hydrocarbons. For the systems under investigation the following kinetic expression is valid:

ln

[reactant]t o [reactant]t

- k 3 t - t o

[reference]t o k1 ln k2 [reference]t

(4)

where [reactant]t0, [reactant]t, [reference]t0 and [reference]t are the initial concentration of the aromatic compound and the concentration at time t, respectively the initial concentration of reference hydrocarbon and the concentration at time t. Hence plots of ln([reactant]t0/[reactant]t) – k3(t-t0) against ln([reference]t0/[reference]t) should yield a straight line of slope k1/k2 with zero intercept. Results The experimental kinetic data plotted in Figure 1 according to eq (4) are preliminary and are presently based on only single experiment and one reference hydrocarbon for each compound. The plots show reasonable linearity considering that the reactions are fairly slow and also the difficulties which arise in the infrared spectral subtraction procedures for these aromatic hydrocarbons. Table 1 shows the preliminary rate coefficients extracted from the plots in Figure 1 for the reaction of OH with the four nitrophenols investigated. A rate constant of 8.52 × 10-12 cm3 s-1 was used for the reaction of OH with the reference ethene (Atkinson and Arey, 2003). Photolysis of methylated nitrophenols is a potential source of OH radicals and thus also a potential source of error in the determination of the correction for photolysis in the kinetic analysis. To obtain the true values of the photolysis rates for the compounds it was necessary to perform separate experiments in which the OH radicals were scavenged. Using isoprene as scavenger it was possible to calculate the OH radical concentration and thus calculate the overestimation in the measured photolysis rate due to reaction with OH. The calculated concentrations of OH radicals produced in the photolysis of the various nitrophenols are presented in Table 2.

I. Bejan et al.

158 Table 1.

Preliminary rate constants for the reactions of OH with four i-methyl-2nitrophenols (i = 3, 4, 5, 6). The literature values (a) are computer estimates.

compound

reference

k1/k2

k1 (10 *cm3s-1)

k (literature)a (10-12*cm3s-1)

3-methyl-2nitrophenol

ethene

0.433±0.019

3.7±0.162

11.2

4-methyl-2nitrophenol

ethene

0.412±0.019

3.5±0.166

5.38

5-methyl-2nitrophenol

ethene

0.671±0.037

5.7±0.313

11.2

6-methyl-2nitrophenol

ethene

0.317±0.019

2.7±0.167

-

Table 2.

-12

Photolysis rates and the steady-state OH radical concentration estimated during the photolysis of all four i-methyl-2-nitrophenols (i = 3, 4, 5, 6) investigated. (a) values for the photolysis rate obtained in the 1080 L quartz glass reactor.

Compound

3-methyl-2nitrophenol

4-methyl-2nitrophenol

5-methyl-2nitrophenol

6-methyl-2nitrophenol

photolysis rate(a) (s-1)

1.67±0.11 x10-4

8.86±1.07x10-5

1.07±0.14x10-4

7.81±1.11x10-5

OH concentration (cm-3)

~3x105

~2x105

~3x105

~2x105

FT-IR Kinetic Study on the Gas-Phase Reactions of the OH Radical

Figure 1.

159

Kinetic data plots for the reactions of methylated-2-nitrophenols with OH radicals using ethene as reference hydrocarbon: (Ŷ) 3-methyl-2-nitrophenol (3M2N), (Ÿ) 4-methyl-2-nitrophenol (4M2N), (i) 5-methyl-2-nitrophenol (5M2N), (x) 6-methyl-2-nitrophenol. (The offsets have been introduced for clarity).

The magnitude of the photolysis rates for methylated nitrophenols under atmospheric conditions can be roughly estimated from the experiments using the superactinic lamps. This is accomplished by adjusting the values obtained in the reactor and presented in Table 2, with a factor comprised of the ratio of NO2 photolysis measured in the atmosphere and that measured in the reactor. For example, a factor of JNO2 (atmosphere)/ JNO2 (in the reactor) = (8.5 ± 0.5) x 10-3 s-1/ (2.0 ± 0.2) x 10-3 s-1 = 4.25 is obtained when using an atmospheric noontime photolysis frequency of NO2 typical for clear sky conditions at a latitude of 40o N for July 1 (Klotz et al., 1997). The loss processes (wall deposition and photolysis) make a large contribution of around 50-55% to the measured total loss of most of the compounds investigated. In the case of 5-methyl-2-nitrophenol the contribution was between 29 - 33%. This lower contribution from photolysis and wall loss to the overall loss for this compound can be explained by the higher reactivity of 5-methyl-2-nitrophenol towards OH radical compared to the other nitrocresols. The losses due solely to photolysis were approximately 23-27% in all cases.

I. Bejan et al.

160 Remarks and conclusions

This work represents the first reported experimental kinetic study of the reaction of OH radicals with methylated nitrophenols, therefore, a comparison with other experimentally obtained values is not possible. Because of the difficulties in handling the compounds the measured rate constants should be used with caution. The measured values presented here need verification by further experiments and also a reference hydrocarbon other than ethene. Validation by an independent absolute kinetic method is also desirable. Up to now there has only been a theoretical estimation of the rate constants for the atmospheric gas-phase reaction of these compounds with OH radicals (Meylan and Howard, 1993). According to the estimate both 3M2N and 5M2N should have the same rate constant of 1.12 × 10-11 cm3 s-1 and 4M2N a value of 5.4 × 10-12 cm3 s-1. 6M2N was not estimated. These results of the present work are in disagreement with the estimated values; all of the measured values are lower than the estimates with the differences being factors of 2 and 3 in the cases of 5M2N and 3M2N, respectively. The measured rate coefficients have values which are larger than that of OH + 2nitrophenol (Atkinson et al. 1992) and lower than those of OH + cresol isomers (Calvert et al. 2002). This is in line with reactivity expectations. The rate coefficient for all four nitrocresols is around 10-12 cm3 s-1, however, the preliminary results appear to indicate an influence of the position of the methyl group on the aromatic ring on the rate coefficient. Based on established chemical activating/deactivating effects for ring substituents (discussed below), the rate for OH + 5M2N would be predicted to have a higher rate than the others and OH + 6M2N should have the lowest rate constant, 3M2N and 4M2N would be expected to have fairly similar rate constants. This is roughly what is observed experimentally. The reaction of phenolic compounds with OH radicals occurs predominantly by electrophilic addition of OH to the aromatic ring (Atkinson, 2000). NO2 is a meta director because electrophilic attack at the ortho and para positions generates an especially poor resonance structure and the CH3 group is and ortho-para director. OH has a strong electron donating resonance effect that activates the ring. For all four investigated compounds NO2 and OH are fixed on adjacent positions on the aromatic ring. Thus the position of CH3 substituent has a profound affect on the reactivity of aromatic compounds with OH radicals, as can be seen in Figure 3. OH

OH NO2

*° *°

CH3

OH NO2

* °

°

NO2

*° H3C

OH H3C

NO2 °

°

*°

*

5M2N

6M2N

CH3 3M2N

Figure 3.

4M2N

The nitrocresol isomers considered in this study and the sites that are activated by the hydroxyl group (*) and methyl group (°) toward electrophilic addition.

The OH and CH3 entities both donate electron density to the aromatic ring and activate the ortho and para positions toward electrophilic addition. The directing + I effect of the methyl group is considerably less than the + E effect of OH, but nonetheless the CH3 group

FT-IR Kinetic Study on the Gas-Phase Reactions of the OH Radical

161

still plays an important role. OH activation and NO2 deactivation effects are in the 4th and 6th positions for all compounds, so additional activation from CH3 becomes important. The faster rate coefficient for the reaction of 5M2N compared with those of the other iM2Ns (i=3,4,6) can be ascribed to enhanced activation of the 4th and 6th ring positions by both the methyl and OH groups (ortho- by CH3, and both ortho- and para- by OH). A similar effect is present in 3M2N, but the + I effect of CH3 is somewhat suppressed by the – E effect of the NO2 group. Within the experimental error limits, the rate coefficients for 4M2N and 6M2N are almost the same. This is expected because the number of possible addition sites for the OH radical is the same for both compounds. Further studies are in progress to validate the rate coefficients presented here and also to determine both the photolysis and OH reaction products. References Atkinson, R. and J. Arey, Atmospheric Degradation of Volatile Organic Compounds, Chem. Rev. 103 (2003) 4605-4638. Atkinson, R., Atmospheric chemistry of VOCs and NOx, Atmos. Environ. 34 (2000) 2063-2101. Atkinson, R., S. Aschmann, and J. Arey, Reaction of OH and NO3 radicals with phenol, cresol, and 2nitrophenol at 296 ± 2 K, Environ. Sci. Technol., 26 (1992), 1397-1403 Barnes, I., K. H. Becker and N. Mihalopoulos, An FT-IR Product Study of the Photo-oxidation of Dimethyl Disulphide, J. Atmos. Chem., 18 (1994) 267-289 Bejan, I., I. Barnes, R. Olariu and R. Mocanu, Secondary organic aerosol formation from the photolysis of nitrophenols and nitrocresols, Poster presented at European Geosciences Union, 1st General Assembly, Nice, France, 25-30 April 2004 Belloli, R., B. Barletta, E. Bolzacchini, S. Meinardi, M. Orlandi and B. Rindone, Determination of toxic nitrophenols in the atmosphere by high-performance liquid chromatography, J. Chromatogr. A, 846 (1999), 277-281. Calvert, J., R. Atkinson, K.H. Becker, R. Kamens, J. Seinfeld, T. Wallington and G. Yarwood, The Mechanisms of the Atmospheric Oxidation of Aromatic Hydrocarbons, Oxford University Press, 2002. Derwent, R. G., M. E. Jenkin, S. M. Saunders and M. J. Pilling, Photochemical ozone creation potentials for organic compounds in Northwest Europe calculated with a master chemical mechanism, Atmos. Environ. 32, 14/15, (1998) 2429-2441. Derwent, R. G., M. E. Jenkin and S. M. Saunders, Photochemical ozone creation potentials for a large number of reactive hydrocarbons under european conditions, Atmos. Environ. 30, 2, (1996) 181-199. Forstner, H.J.L., R.C. Flagan and J.H. Seinfeld, Secondary organic aerosol from the photo-oxidation of aromatic hydrocarbon: Molecular composition, Environ. Sci. Techn 31, (1997) 1346-1358. Geissler, A. and H. F. Scholer, Gas chromatographic determination of phenol, methylphenols, chlorophenols, nitrophenols and nitroquinones in water. Water Research 28 (1994) 2047–2053. Grosjean, D., Atmospheric fate of toxic aromatic compound, Sci. Total Environ., 100 (1991), 367-414. Harrison, M. A. J., S. Barra, D. Borghesi, D. Vione, C. Arsene and R. I. Olariu, Nitrated phenols in the atmosphere: a review, Atmos. Environ., 39 (2005) 231-248 Herterich, R., Gas chromatographic determination of nitrophenols in atmospheric liquid water and airborne particulates. J. Chromatography, 549 (1991) 313-324. Hurley, M.D., O. Sokolov, T.J. Wallington, H. Takekawa, M. Karasawa, B. Klotz, I. Barnes and K.H. Becker, Organic Aerosol Formation During the Atmospheric Degradation of Toluene, Environ. Sci. Techn., 35 (2001) 1358-1366. Kawai, A., S. Goto, Y. Matsumoto and H. Matsushita, Mutagenicity of aliphatic and aromatic nitro compounds industrial materials and related compounds, Jap. J. Ind. Health, 29 (1987), 34-54.

162

I. Bejan et al.

Kawamura K. and I. R. Kaplan, Biogenic and anthropogenic organic compounds in rain and snow samples collected in southern California, Atmos. Environ. 20 (1986) 115–124. Klotz, B., I. Barnes, K.H. Becker and B. T. Golding, Atmospheric chemistry of benzene oxide/oxepin, J.Chem. Soc., Faraday Trans., 93 (1997) 1507-1516 Lüttke, J. and K. Levsen, Phase partitioning of phenol and nitrophenols in clouds Atmos. Environ., 31 (1997), 2649-2655. Meylan, W. M. and P. H. Howard, Computer estimation of the atmospheric gas-phase reaction rate of organic compounds with hydroxyl radicals and ozone, Chemosphere, 26, 12 (1993) 2293-2299. Nojima, K., K. Fukaya, S. Fukui and S. Kanno, The formation of nitrophenol and nitrobenzene by the photochemical reaction of benzene in the presence of nitrogen monoxide, Chemosphere, 2 (1975), 77-82. Odum, J.R., T. Hoffmann, F. Bowman, D. Collins, R.C. Flagan and J.H. Seinfeld, Gas/particle partitioning and secondary organic aerosol yields, Environ. Sci. Techn. 30 (1996) 2580-2585. Olariu, R. I., B. Klotz, I. Barnes, K. H. Becker and R. Mocanu, FT-IR study of the ring- retaining products from the reaction of OH radicals with phenol, o-, m- and p-cresol, Atmos. Environ., 36 (2002) 3685-3697 Tremp, J., P. Mattrel, S. Fingler, W. Giger, Phenol and nitrophenols as tropospheric pollutants: emissions from automobile exhausts and phase transfer in the atmosphere. Water Air Soil Pollut., 68 (1993), 113-123. Voznakova, Z., J. Podehradska, and M. Kohlickova, Determination of nitrophenol in soil, Chemosphere, 33, 2 (1996), 285-291 Winzor, F. L., The colouring matters of Drosera Whittakeri. Part III. The synthesis of hydroxydroserone. J. Chem. Soc. Lond. (1935) 336-338

Atmospheric Fate of Unsaturated Ethers A. Mellouki LCSR/CNRS, 1C Avenue de la recherche scientifique, 45071 Orléans cedex 02, France Key Words: Unsaturated Ethers, Tropospheric degradation, Kinetics, OH radicals, NO3 radicals, Ozone

Introduction Oxygenated volatile organic compounds form a major component of the trace gases found in the troposphere (Singh et al., 2001). They are emitted directly into the atmosphere from biogenic sources, solvent and fuel additives use and are also formed in the tropospheric oxidation of all hydrocarbons. A number of oxygenated compounds are used as alternatives to aromatic and halocarbon solvents both in terms of toxicity problems and in the case of aromatic compounds as a means of reducing the levels of oxidant formation in the troposphere. It is apparent that oxygenated compounds will thus play an increasing role in determining the oxidising capacity of the troposphere both on a regional and global scale. The degradation of oxygenated volatile organic compounds in the troposphere leads to the production of a range of secondary pollutants such as ozone, nitrates and secondary organic aerosols. Ozone formation is of particular concern since it is known to have adverse effects on human health, vegetation and a variety of materials, and it is also a greenhouse gas. Indeed, strategies for the control of emissions of volatile organic compounds have been based on their individual contribution to the photochemical formation of ozone (Derwent et al., 1996). Information on the fate of oxygenated compounds in the troposphere is not only important in terms of oxidant formation, but oxidation of simple oxygenated compounds often leads to the generation of poly-functionalized species. These compounds can be of low reactivity in the gas phase and thus their relatively high solubility and low vapour pressures may lead to uptake in cloud and fog droplets or to formation of secondary organic aerosols. Hence, anthropogenic emissions of oxygenates may impact on wet deposition and also on climate by increasing the levels of tropospheric aerosol. Because of their high reactivity, vinyl ethers can be used in a variety of syntheses: e.g. polymerizations, addition, and electrocyclic reactions. They are used in different industries, particularly as solvents, motor oil additives, for the manufacturing of coatings or as intermediates for the synthesis of flavours, fragrances and pharmaceuticals. This broad utility has made many applications possible, and is expected to create even more applications in the future (http://www.basf.de). In the atmosphere, the degradation of unsaturated oxygenates is controlled mainly by chemical reaction with OH and NO3 radicals and O3 while photolysis plays a negligible role. A number of studies have been conducted recently, within an EU project to investigate the atmospheric fate of the unsaturated ethers (Georges et al., 2004). In this contribution, some of the experimental data obtained on the gas phase processes are presented. This includes kinetic and mechanistic data for the reactions of OH, O3 and NO3 with a series of vinyl ethers (methyl vinyl ether (MVE) CH3OCH=CH2, ethyl vinyl ether (EVE) C2H5OCH=CH2, propyl vinyl ether (PVE) n-C3H7OCH=CH2, butyl vinyl ether (BVE) n-C4H9OCH=CH2, iso-butyl vinyl ether (IBVE) i-C4H9OCH=CH2 and tert-butyl vinyl ether (TBVE) t-C4H9OCH=CH2). 163 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 163–169. © 2006 Springer. Printed in the Netherlands.

A. Mellouki

164 OH-initiated oxidation of vinyl ethers

The rate constants of the reactions of OH with the vinyl ethers have been measured in the temperature and pressure ranges 230-373 K and 30-300 Torr using the pulsed laser photolysis-laser induced fluorescence (PLP – LIF) method and also the relative kinetic method at 760 Torr and 298 K. The measured rate coefficients in the temperature range 230372 K are shown in the conventional Arrhenius form (k = Ae-Ea/RT) in Figure 1. The data obtained at room temperature as well as the Arrhenius parameters are summarised in Table 1. The Arrhenius parameters derived from the temperature dependence study show a negative temperature dependence for all measured rate constants which supports a mechanism proceeding mainly by addition of OH to the double bond of the vinyl ether.

Figure 1.

Rate constants for the reactions of OH with the studied vinyl ethers in the temperature range 230-373 K plotted according to the Arrhenius expression. The solid lines represent the Arrhenius parameter least-squares fit to the individual data points. The error bars of the individual points are 2V and do not include systematic errors (Thiault et al., 2002 and 2005).

The rate constant values of the OH reaction with vinyl ethers are higher than those with simple ethers (ROR’, with R, R’ = CH3, C2H5, C3H7, C4H9), in the ranges (4 – 11) × 10-11 and (0.3 – 3) × 10-11 cm3 molecule-1 s-1, respectively. This is in line with an OH-addition mechanism to the double bond as in the reaction of OH with alkenes. However, the values of the OH reactions with vinyl ethers, as shown in Table 2, are up to three times higher than those of OH with the corresponding alkenes. This would suggest that –OR groups in the vinyl ethers activate the double bond by electron donation more than alkyl groups –R do. In addition, while the reactivity of OH with alkenes does not seem to be influenced by the nature of the R group (k(OH+alkene) | 3×10-11 cm3 molecule-1 s-1) with R = CH3 to C4H9, that of OH

Atmospheric Fate of Unsaturated Ethers

165

with vinyl ethers increases from 4×10-11 for CH3OCH=CH2 (Perry et al., 1977) to 1.1×10-10 cm3 molecule-1 s-1 for C3H7OCH=CH2 and C4H9OCH=CH2. Table 1.

Summary of the measured rate constants at 298 K obtained by absolute and relative methods and the Arrhenius parameters in the temperature range 230372 K. a)

Vinyl ether

a)

a)

A u 10

11

b)

k298 K u 1010

a)

k298 K u 1010

E/R Absolute

Relative

MVE, CH3OCH=CH2

-

-

-

0.45 r 0.07

EVE, C2H5OCH=CH2

1.55 r 0.25

-(445 r 13)

0.68 r 0.07

0.72 r 0.08

PVE, C3H7OCH=CH2

0.93 r 0.06

-(708 r 20)

1.0 r 0.1

1.1 r 0.1

BVE, n-C4H9OCH=CH2

1.48 r 0.22

-(572 r 42)

1.0 r 0.1

1.1 r 0.1

IBVE, i-C4H9OCH=CH2

1.57 r 0.11

-(567 r 20)

1.1 r 0.1

1.1 r 0.1

TBVE, t-C4H9OCH=CH2

1.70 r 0.15

-(549 r 25)

1.1 r 0.1

1.1 r 0.1

in cm3 molecule-1 s-1 ; b) in K.

Table 2.

Comparison of the rate constant values for the reaction of OH radicals with vinyl ethers and alkenes at 298 K (k in cm3 molecule-1 s-1).

Alkyl group (R)

Alkenes (RCH=CH2)

Vinyl Ethers (ROCH=CH2)

CH3

2.63 x 10-11

3.4 x 10-11

C2 H5

3.14 x 10-11

7 x 10-11

n-C3H7

3.14 x 10-11

10 x 10-11

n-C4H9

3.7 x 10-11

11 x 10-11

i-C4H9

3.8 x 10-11

11 x 10-11

t-C4H9

2.8 x 10-11

11.2 x 10-11

As shown in Table 3, the structure reactivity relationships (SAR) developed by Atkinson et al. does not apply to the vinyl ethers since the calculated rate constant values are lower than the measured ones for the vinyl ethers studied in this work. The main observed products of the OH-initiated oxidation of vinyl ethers in the presence of NOx, were formaldehyde and formates (with molar yields of | 80 % for each).

A. Mellouki

166

The observed products can be explained by a mechanism proceeding mainly by OH-radical addition to the carbon-carbon double bond: ROCH=CH2 + OH (+O2 + NO + M) o ROCH(OH)CH2O° + NO2

(1a)

ROCH=CH2 + OH (+O2 + NO + M) o ROCH(O°)CH2OH + NO2

(1b)

The hydroxyalkoxy radicals ROCH(OH)CH2O° and ROCH(O°)CH2OH decompose, react with O2, or isomerise. The reactions of these two radicals with O2 may lead to hydroxycarbonyl compounds (ROCH(OH)CHO and ROC(O)CH2OH) while their decomposition will lead to the corresponding formate and formaldehyde: ROCH(OH)CH2O° (+M) o ROC°HOH + HCHO ROC°HOH + O2 o ROC(O)H + HO2 and ROCH(O)CH2OH (+M) o ROC(O)H + CH2OH CH2OH + O2 o HCHO + HO2 Table 3.

Comparison between the measured rate constant values for the reaction of OH with vinyl ethers and the calculated values using the SAR of Atkinson et al. (Atkinson, 1987; Kwok and Atkinson, 1995).

Vinyl ether

kcalculated

kmeasured

(cm3 molecule-1 s-1)

(cm3 molecule-1 s-1)

CH3OCH=CH2

3.5 x 10-11

3.4 x 10-11 (Perry et al., 1977)

C2H5OCH=CH2

4.2 x 10-11

7 x 10-11

n-C3H7OCH=CH2

4.5 x 10-11

10 x 10-11

n-C4H9OCH=CH2

4.7 x 10-11

11 x 10-11

i-C4H9OCH=CH2

4.7 x 10-11

11 x 10-11

t-C4H9OCH=CH2

3.5 x 10-11

11 x 10-11

The detection of high yields of formates and formaldehyde indicates that the decomposition of the two hydroxyalkoxy radicals dominates over the other processes (reaction with O2 and isomerisation). The other possible mechanism is H-atom abstraction from the R groups which produces vinyl esters such as R’C(O)OCH=CH2 and HC(O)OCH=CH2 when the abstraction is from the –CH2- group attached to –O- entity: R’CH2OCH=CH2 + OH (+O2 + NO + M) o R’CH(O°)OCH=CH2 + NO2 R’CH(O°)OCH=CH2 (+O2) o R’C(O)OCH=CH2 R’CH(O°)OCH=CH2 (+M +O2) o HCHO + H(O)COCH=CH2

Atmospheric Fate of Unsaturated Ethers

167

An H-abstraction from the R’ group will lead to other species, however, the only products observed were formaldehyde and formates which indicates that the OH-radical oxidation of vinyl ethers proceeds essentially by OH addition to the carbon-carbon double bond. O3 Reactions with Vinyl Ethers The measured values for the rate constants of the reaction of O3 with vinyl ethers are presented in Table 4. The are in the range (2-5)×10-16 cm3 molecule-1 s-1, i.e. they are higher than those for simple alkenes again indicating that the presence of the O atom also in this case enhances the reactivity with the C=C double bond. It would appear that the nature of the R group in the vinyl ethers (ROCH=CH2) also has an effect on the reactivity. As in the OH-initiated oxidation of vinyl ether, the reactions of O3 with the same vinyl ethers lead mainly to formation of formates (molar yields of (70-90)%) and formaldehyde (molar yields of (20-10)%). Figure 2 shows an example of the concentration-time profiles of the reactants and products for the reaction of O3 with ethyl vinyl ether (C2H5OCH=CH2). For methyl vinyl ether, hydroperoxymethyl formate and methanol were also observed as products (Klotz et al., 2004). Table 4.

O3 + vinyl ethers, summary of the measured rate constants at 298 ± 2 K. Vinyl Ethers

k (cm3 molecule-1 s-1)

EVE, C2H5OCH=CH2

(2.0 r 0.2) u 10-16

PVE, C3H7OCH=CH2

(2.4 r 0.4) u 10-16

BVE, n-C4H9OCH=CH2

(2.9 r 0.2) u 10-16

IBVE, (CH3)2CHCH2OCH=CH2

(3.1 r 0.2) u 10-16

TBVE, (CH3)3COCH=CH2

(5.0 r 0.5) u 10-16

As for the reactions of O3 with olefins, the reaction of O3 with ROCH=CH2 proceeds by initial addition of O3 to the carbon-carbon double bond, followed by decomposition of the ozonide to form formaldehyde and a formate as stable compounds in addition to two Criegee intermediates: O3 + ROCH=CH2 ĺ ROC°HOO° + HCHO O3 + ROCH=CH2 ĺ ROC(O)H + HC°HOO° The energy rich Criegee intermediates ROC°HOO° and HC°HOO° may thermally stabilize, rearrange, react or decompose to form other stable compounds. Laboratory studies have shown that the O3 initiated oxidation of vinyl ethers may lead to the formation of secondary organic aerosols. The reported mass based formation yields are in the range (0.5-10) % depending on the starting organic compound (Klotz et al., 2004; Georges et al., 2004).

A. Mellouki

168

EVE Ethyl Formate HCHO Ozone

EVE concentration (ppb)

800

200

700

100 600

0

500

0

Figure 2.

Products and ozone concentrations (ppb)

300

900

5

10 15 Reaction time (min)

20

25

Concentrations of products and reactants versus reaction time in the O3 initiated oxidation of C2H5OCH=CH2 in air (Thiault et al., 2002).

NO3 Reaction with vinyl ethers Kinetic and mechanistic investigations on the NO3 radical oxidation of three vinyl ethers, methyl, propyl and butyl vinyl ethers have been performed in smog chambers (Klotz et al., 2004, Georges et al., 2004). Analyses were performed using an in situ FTIR device coupled to a multi-reflection White-cell. Rate constants for the oxidation of methyl and propyl vinyl ethers with NO3 were determined using the relative rate technique and/or by a “pseudoabsolute” rate technique. Rate constants of (4 r 2) ×10-13 and (1.3 r 0.2) ×10-12 cm3 molecule1 -1 s were obtained for the NO3 oxidation of methyl and propyl vinyl ethers, respectively. Mechanistic experiments also have been performed for methyl, propyl and butyl vinyl ethers. Main oxidation products are formaldehyde, a formate (methyl, propyl or butyl), nitrates and peroxynitrates (Figure 3). Formaldehyde and the formates have been quantified and their formation yields estimated to be around 50 %. The reaction of NO3 with unsaturated ethers proceeds by addition to the >C=C< double bond. Both steric and inductive effects are known to influence the reactivity of NO3 with unsaturated alcohols, e.g. 2-methyl-3-butene-2-ol is slightly less reactive than 1-butene. Unsaturated ethers are more reactive than unsaturated alcohols towards NO3. Comparison of the measured rate constants shows that, similar to the reaction with O3, the presence of the carbonyl group decreases the reactivity of NO3 toward the >C=C< double bond. The reaction of NO3 with unsaturated oxygenates proceeds by addition to either one of the carbon atoms of >C=C< double bond, preferentially to the most substituted radical. In the presence of NOx, the reaction may ultimately lead to organic nitrates among the first generation products.

Atmospheric Fate of Unsaturated Ethers

169

3

PVE + NO3 PVE

2.5

Propyl Formate

HCHO

Concentration (ppm)

N2O5

HNO3

2

NO2 Sum Nitrates (arbitrary units)

1.5

1

0.5

0 0

500

1000

1500

2000

2500

3000

3500

Time (s)

Figure 3.

Concentration profiles for an experiment of NO3-induced oxidation of propyl vinyl ether (Georges et al., 2004).

Atmospheric Implications In the atmosphere, unsaturated ethers are oxidised in the gas phase by reaction with OH, NO3 and O3. The tropospheric lifetimes of unsaturated ethers are typically of few hours and oxidation results mainly in the production of aldehydes and formates which are less reactive than their parent compounds. It has been found that the reactions of ozone with vinyl ethers lead to the formation of secondary organic aerosols under laboratory conditions. This aerosol formation warrants further investigation in order to assess its importance under real atmospheric conditions. References Atkinson R.; A structure activity relationship for the estimation of rate constants for the gas-phase reactions of OH radicals with organic compounds, Int. J. Chem. Kinet. 19 (1987) 799-828. Derwent, R. G.; Jenkin, M. E.; Saunders, S. M. Atmos. Environ. 30 (1996) 181. Georges, C. et al., Multiphase chemistry of Oxygenated Species in the Troposphere – MOST, EVK2-CT-200100114 (2002-2004). Klotz B., I. Barnes and T. Imamura; Product study of the gas-phase reactions of O3, OH and NO3 radicals with methyl vinyl ether, Phys. Chem. Chem. Phys. 6 (2004) 1725-1734. Kwok E. S. C. and R. Atkinson; estimation of hydroxyl radical reaction rate constants for gas phase organic compounds using a structure-reactivity relationship: an update, Atmos. Environ. 29 (1995) 1685-1695. Perry R.A., R. Atkinson, J. N. Pitts; Rate Constants for the Reaction of OH Radicals with Dimethyl Ether and Vinyl Methyl Ether over the Temperature Range 299-427oK, J. Chem. Phys., 67 (1977) 611. Singh, H.; Chen, Y.; Standt, A.; Jacob, D.; Blake, D.; Heikes, B.; Snow, J. Nature 410 (2001) 1078-1081. Thiault G., R. Thévenet, A. Mellouki, G. Le Bras; OH and O3-initiated oxidation of ethyl vinyl ether Phys. Chem. Chem. Phys. 4 (2002) 613-619. Thiault G., A. Mellouki, G. Le Bras ; Rate constants for the OH radical reaction with n-propyl, n-butyl, iso-butyl and tert-butyl vinyl ethers to be submitted (2005).

Atmospheric Oxidation of the Chlorinated Solvents, 1,1,1-Trichloroethane, Trichloroethene and Tetrachloroethene Lorraine Nolan, Anne-Laure Guihur, Marcus Manning, and Howard Sidebottom Department of Chemistry, University College Dublin, Belfield, Dublin 4, Ireland Key Words: Chlorinated solvents, Chloroalkoxy radicals, Photolysis, Troposphere, Hydroxyl radical

Introduction Chlorinated solvents are widely used in metal degreasing, dry cleaning, paint stripping, extraction of pharmaceuticals and foodstuffs, and electronic circuit board production (Sidebottom and Franklin, 1996). Emission data for trichloroethene (CHCl=CCl2) and tetrachloroethene (CCl2=CCl2) show that emissions of these compounds are declining steadily, largely as a result of constant improvements in the efficiency with which they are being used and recycled. While 1,1,1-trichloroethane (CH3CCl3) was formerly widely used as a solvent, it is now strictly regulated as an ozone-depleting compound under the Montreal Agreement. Despite the decline in emissions of these chlorinated solvents their atmospheric fate and impact is still of concern. Chloral (CCl3C(O)H) is the major product from the OH radical initiated oxidation of CH3CCl3 (Nelson et al., 1990, Platz et al., 1995). The available data show that the tropospheric lifetime of CCl3C(O)H is of the order of a few hours and is predominantly controlled by photolysis (Talukdar et al., 2001, Wenger et al., 2004). It has been shown that a chlorine atom is produced in the primary photolysis process at 308 nm with a quantum yield close to unity and it was assumed that the HC(O)CCl2 radical, which is also formed in the photolysis, generates phosgene (Cl2C(O)) in the presence of O2 (Talukdar et al., 2001). Recently, the photolysis of chloral under atmospheric conditions has been investigated at the large outdoor European photoreactor in Valencia, Spain (Wenger et al., 2004). The photolytic lifetime of chloral was shown to be 4.5-6 hours under conditions appropriate to the solar flux during summer months and the quantum efficiency of photolysis under atmospheric conditions was determined to be 1.00 ± 0.05. The reported atmospheric photolysis product yield data are consistent with a mechanism in which the primary process generates a Cl atom and a HC(O)CCl2 radical. The HC(O)CCl2 radical forms the corresponding alkoxy radical 3HC(O)CCl2O which decomposes by both C-C (~85%) and C-Cl (~15%) bond fission. This work is concerned with reactions of the alkoxy radicals HC(O)CCl2O and ClC(O)CCl2O under atmospheric conditions in order to determine the relative importance of the carbon-carbon and carbon-chlorine bond fission pathways for their degradation. The OH radical initiated oxidations of the chloroethenes CHCl=CCl2 and CCl2=CCl2 in the presence of a Cl atom scavenger have been reported to give Cl2C(O) as a major product with yields of around 50% for both alkenes (Tuazon et al., 1988). The relatively low yields of Cl2C(O) and the lack of HClC(O) as a product in the oxidation of CHCl=CCl2 suggests that the energy-rich alkene adduct, formed in the reactions, may eliminate HCl rather than add molecular oxygen. The resulting radicals will lead to the alkoxy radicals HC(O)CCl2O and ClC(O)CCl2O. Hence the breakdown pathways of these radicals may be of importance in understanding the atmospheric oxidation of tri-and tetrachloroethene. 171 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 171–179. © 2006 Springer. Printed in the Netherlands.

L. Nolan et al.

172 Experimental

Product distribution studies in the laboratory were carried out at 298 ± 2K and atmospheric pressure in a FEP Teflon chamber with a volume of approximately 50 L. The temperature of the reaction chamber was thermostatically controlled and electric fans positioned below the chamber helped maintain a constant temperature. The reaction chamber was surrounded by banks of fluorescent lamps, which provided sources of radiation at 254 nm (Philips germicidal lamps TUV, 15 W) and in the 300 to 450 nm wavelength region with a maximum intensity around 365.5 nm (Philips black lamps TL 20/08). The reaction chamber was approximately half-filled by a diluent gas, typically zero-grade air, via Teflon tubing using calibrated flow meters. Measured pressures (MKS Baratron capacitance manometers) of the reactants were then expanded into the evacuated vacuum line and isolated in a calibrated mixing bulb, from which they were swept into the reaction chamber by a stream of zero-grade air. Hydrogen peroxide solution was injected directly into the chamber. When all the reactants had been added to the chamber it was subsequently filled to capacity at a pressure of 1 atmosphere and kept in the dark for 1 hour to allow complete mixing of the reactants. A homogeneous reaction mixture was confirmed by consistent, reproducible gas chromatographic analysis. Hydroxyl radicals were generated by the photolysis of hydrogen peroxide. H2O2 + hȞ (Ȝ = 254 nm)

ĺ

OH + OH

(1)

Molecular chlorine was photolysed to produce chlorine atoms. Cl2 + hȞ (Ȝ > 300 nm)

ĺ

Cl + Cl

(2)

Quantitative analyses of the reaction mixtures were carried out by gas chromatography (Shimadzu GC 14A coupled with a flame ionisation detection), and Fourier transform infrared spectroscopy (Mattson Research Series 1 fitted with a deuterated triglycine sulfate (DTGS) detector). The spectrometer was equipped with an evacuable 5 L Teflon–coated Wilks cell, containing a multipass White mirror arrangement (pathlength 7-10 m) to facilitate the collection of infrared spectra. Spectra were recorded over the range 400-4000 cm-1 and with a resolution of 2 cm-1. Each resulting spectrum was derived from the co-addition of 20 scans under these conditions. Authentic samples of reactants and products were used to calibrate reference spectra. The reaction chamber employed for product studies at the EUPHORE facility consisted of a half-spherical FEP Teflon chamber with a volume of around 195,000 L mounted on aluminium floor panels covered with FEP Teflon foil. The FEP Teflon had a transmission of more than 80% over the wavelength range 280-640 nm. A hydraulically operated steel housing surrounded the chamber and allowed the exposure time to sunlight to be controlled. Heating of the reaction mixtures above ambient temperature was compensated with cooling by a refrigeration system housed in the chamber floor. Typically, the chamber temperature was in the range 300 ± 5 K. The chamber was filled with air from the air purification system consisting of absorption dryers connected to charcoal scrubbers. Two mixing fans housed in the chamber were used to ensure homogenous reaction mixtures. Measured volumes of gaseous and liquid samples were introduced into the covered reaction chamber through Teflon tubing by a stream of purified air. Losses of reactants and products occurred by leakage from the chamber, which was maintained at several Torr above ambient pressure. The leak rate was determined by adding an unreactive tracer gas, SF6 to the chamber. The loss of reactants and formation of products was monitored with a Nicolet

Atmospheric Oxidation of the Chlorinated Solvents

173

Magma 550 FTIR spectrometer equipped with a mercury cadmium telluride (MCT) detector positioned on a platform beneath the chamber. Infrared spectra were obtained by in situ longpath absorption, and were derived from the co-addition of 900 scans recorded by using a path length of 553.5 m over the range 400-4000 cm-1 with a resolution of 1 cm-1. The reactants were quantified by using calibrated reference spectra. Results and discussion The photolysis of chloral (50-100 ppm) in air at Ȝ > 300 nm was carried out in the laboratory in the presence of excess isopropanol (250-500 ppm) as a chlorine atom trap. The major products from the reaction were Cl2C(O) and CO with yields of 0.87 ± 0.03 and 1.01 ± 0.03 respectively. The loss of the added isopropanol (1.28 ± 0.06) and consequent formation of acetone (1.30 ± 0.05) provide a measure of the yield of chlorine atoms in the system. Cl + (CH3)2CHOH

ĺ

(CH3)2COH + HCl

(3)

(CH3)2COH + O2

ĺ

HO2 + (CH3)2C(O)

(4)

A similar series of photolysis experiments was carried out at the European outdoor reaction chamber with initial concentrations of chloral and isopropanol in the range 0.4-1.2 ppm. The measured yields of Cl2C(O) (0.85 ± 0.01), CO (1.05 ± 0.05), (CH3)2CHOH (-1.39 ± 0.03) and (CH3)2C(O) (1.28 ± 0.03) were in agreement with the experimental uncertainties determined in the laboratory system. Assuming the photolysis of chloral at Ȝ > 300 nm yields a chlorine atom and a HC(O)CCl2 radical (Talukdar et al., 2001), the yield of Cl2C(O) from the photolysis of chloral indicates that the HC(O)CCl2 radical is not quantitatively converted to Cl2C(O). Thus, decomposition of the HC(O)CCl2O radical, generated from HC(O)CCl2, occurs by either carbon-carbon or carbon-chlorine bond fission (Figure 1). According to the mechanism given in Figure 1, C-C bond cleavage of the HC(O)CCl2O radical will generate Cl2C(O) with a yield of ~0.85 and leads to the formation of CO (~0.85) following the reaction of HC(O) with O2. The remaining ~15% of the HC(O)CCl2O radicals undergo C-Cl bond fission to produce chloroglyoxal HC(O)C(O)Cl and a chlorine atom. This proposed mechanism predicts molar yields of Cl2C(O) (~0.85), CO (~0.85), HC(O)C(O)Cl (~0.15) and Cl (~1.15) as compared to the observed yields of Cl2C(O) (~0.85), CO (~1.0) and Cl (~1.35). Chloroglyoxal was not observed among the products, however it is likely that this compound undergoes fairly rapid photolysis. It is proposed that HC(O)C(O)Cl degrades during the relatively long (~5 hours) chloral photolysis experiments to form the additional amounts of CO and Cl atoms observed. It is interesting to note that the yield of phosgene in experiments carried out in the presence of NO was virtually unchanged from the yields observed in the absence of NO. The HC(O)CCl2O radical produced from the reaction of HC(O)CCl2O2 with NO is vibrationally excited. However, the relative importance of the two unimolecular decomposition channels are almost the same as in the decomposition of HC(O)CCl2O radicals generated by the thermoneutral combination of the corresponding peroxy radicals.

L. Nolan et al.

174 CCl3C(O)H + hQ (O~ 300 nm) HC(O)CCl2 + O2 + M

HC(O)CCl2 + Cl HC(O)CCl2O2 + M

X2 / NO

NO2

HC(O)CCl2O ' (C-C)

' (C-Cl) ~15%

~85%

HC(O)C(O)Cl +Cl

HC(O) + Cl2C(O) O2

hQ

HO2 +CO

HC(O) + ClC(O) O2

HO2 + CO Figure 1.

Cl + CO

Mechanism for the photolysis of CCl3C(O)H in air.

In an attempt to further investigate reactions of alkoxy radicals of general structure RC(O)CCl2O, the chlorine atom initiated oxidation of CHCl2C(O)Cl (50-100 ppm) was investigated in air at 298 K. Product distribution studies show Cl2C(O) (0.90 ± 0.03) and ClC(O)C(O)Cl (0.11 ± 0.01) as the major products. Hydrogen atom abstraction from the methyl group in dichloroacetyl chloride leads to the formation of ClC(O)CCl2 radicals which, following addition to molecular oxygen and reaction with another peroxy radical, generate the ClC(O)CCl2O radical (Figure 2). The yields of Cl2C(O) and ClC(O)C(O)Cl indicate that the ClC(O)CCl2O radicals decay both by the C-C and C-Cl bond cleavage in a similar manner to that suggested for the structurally similar radical HC(O)CCl2O. Since the Cl2 photolysis experiments are relatively rapid (~15 minutes) the dicarbonyl ClC(O)C(O)Cl is photochemically stable over the time period of the experiment.

Atmospheric Oxidation of the Chlorinated Solvents

175

Cl2 + hQ (O > 300 nm)

Cl + Cl

Cl +CHCl2C(O)Cl

Cl(O)CCl2 +HCl

ClC(O)CCl2 + O2 + M

ClC(O)CCl2O2 + M

X2

ClC(O)CCl2O ' (C-C) ~90%

ClC(O) + Cl2C(O)

'(C-Cl) ~10%

ClC(O)C(O)Cl + Cl

Cl + CO

Figure 2.

Mechanism for the chlorine atom initiated oxidation of CHCl2C(O)Cl.

The hydroxyl radical initiated oxidations of trichloroethene (50-100 ppm) and tetrachloroethene (50-100 ppm) were investigated in air at 1 atmosphere total pressure and 298 K in the laboratory reactor. Photolysis of hydrogen peroxide (100-500 ppm) at 254 nm was used as the OH radical source. Reactions were carried out both in the absence and presence of dimethyl ether, which was used as a chlorine atom scavenger. The products found from the reactions of OH with CHCl=CCl2 and CCl2=CCl2 in the absence of CH3OCH3 show the presence of CHCl2C(O)Cl and CCl3C(O)Cl respectively, formed in the Cl atom initiated oxidation of the alkenes. Under conditions where sufficient CH3OCH3 had been added to the reaction mixtures so that the yields of acid chlorides were negligible, it was assumed that the Cl atoms generated in the OH reactions were totally scavenged. The major products detected in the OH radical initiated oxidations of CHCl=CCl2 and CCl2=CCl2 for reactions performed in the presence of sufficient CH3OCH3 to sufficiently quench the Cl atom reactions with the chloroalkenes were Cl2C(O) and CO. The yields of Cl2C(O) in both systems were approximately 70% in reasonable agreement with previously reported data (Tuazon et al., 1988). The yield of HClC(O) from the reaction of OH with CHCl=CCl2 was shown to be less than 5%. The OH radical initiated oxidations of the chloroethenes proceed by initial OH addition to the carbon-carbon double bond. All the available evidence suggests that for

L. Nolan et al.

176

unsymmetric alkenes, OH addition to the least substituted carbon atom dominates and hence reaction with CHCl=CCl2 is expected to involve the addition to the =CHCl group (Atkinson, 1994). While it is known at low pressures the hydroxyalkyl radical formed may unimolecularly decay by loss of OH, it is generally assumed that under atmospheric conditions the hydroxyalkyl formed will rapidly add O2 (Atkinson, 1994). Reaction with CHCl=CCl2 and CCl2=CCl2 can potentially lead to a complex series of reactions since the OH radical adds to a carbon atom that is bonded to a chlorine atom. The adduct radical formed by addition of OH to the alkene may eliminate a Cl atom or HCl (Figures 3 and 4).

HOCHCl-CCl2

OH + CHCl=CCl2

' (C-Cl) O / X2

HOCH=CCl2 + Cl

HC(O)CCl2 +HCl

HOCHClCCl2O ' (C-C)

CHCl2C(O)H

O / X2

HC(O)CCl2O

HOCHCl +Cl2C(O) ' (C-C)

O2

HO2 + HClC(O)

HC(O) + Cl2C(O)

' (C-Cl)

HC(O)C(O)Cl +Cl

O2

HO2 + CO

Figure 3.

Mechanism for the hydroxyl radical initiated oxidation of CHCl=CCl2.

Atmospheric Oxidation of the Chlorinated Solvents

177

HOCCl2-CCl2

OH + CCl2=CCl2

' (C-Cl) O / X2

HOCCl=CCl2 + Cl

ClC(O)CCl2 +HCl

HOCHCl2-CCl2O ' (C-C)

CHCl2C(O)Cl

O / X2

ClCC(O)CCl2O

HOCCl2 +Cl2C(O) ' (C-C)

O2

HO2 + Cl2C(O)

ClC(O) + Cl2C(O)

' (C-Cl)

ClC(O)C(O)Cl +Cl

O2

Cl + CO

Figure 4.

Mechanism for the hydroxyl radical initiated oxidation of CCl2=CCl2.

The absence of CHCl2C(O)H and CHCl2C(O)Cl from the reaction products of the OH initiated oxidation of CHCl=CCl2 and CCl2=CCl2 respectively indicates that Cl atom elimination from the OH-chloroalkene adducts is of minor importance in 1 atmosphere of air. Addition of OH radicals to CHCl=CCl2 and CCl2=CCl2 gave Cl2C(O) and CO as the major oxidation products. The relatively low yield of Cl2C(O) in the OH initiated oxidation of CCl2=CCl2 and formation of Cl2C(O) but the absence of HClC(O) in the OH initiated oxidation of CHCl=CCl2 provides evidence that elimination of HCl occurs from the initially formed energy rich adducts in these reactions. Thus, the results are consistent with the formation of HC(O)CCl2 and ClC(O)CCl2 radicals in the reactions of OH with CHCl=CCl2 and CCl2=CCl2 (Figures 3 and 4 respectively). These radicals are identical to the radicals formed by photolysis of chloral and the reaction of Cl atoms with CHCl2C(O)Cl. The absence of HClC(O) and the relatively low yield of Cl2C(O) in the CHCl=CCl2 and CCl2=CCl2 systems suggest that addition of OH to the chloroalkenes leads mainly to loss of HCl rather than addition of the adduct radicals to molecular oxygen.

L. Nolan et al.

178

Rather interestingly further information on the mechanism for the OH radical initiated oxidation of CHCl=CCl2 comes from studies on the reactions of Cl atoms with 2,2,2trichloroethanol in the presence of O2. Reaction of Cl atoms with CCl3CH2OH involves Hatom abstraction from the –CH2- group to give the CCl3CHOH radical. Reaction of this radical with O2 is expected to yield CCl3C(O)H, Figure 5. Formyl chloride and phosgene were also identified as primary oxidation products and it is suggested that these products arise from a 1,2 chlorine atom rearrangement of CCl3CHOH to give the HOCHClCCl2 radical. Reaction of this radical with molecular oxygen followed by self reaction of the peroxy radicals will generate the ȕ–hydroxy radical HOCHClCCl2O. The initial yields of CCl3C(O)H (~0.5), Cl2C(O) (~0.4) and HClC(O) (~0.4) from reactions carried out in air at 1 atmosphere suggests that under these conditions around 50% of the alkyl radicals CCl3CHOH, formed by H-atom abstraction from the parent alcohol, react with O2 to form CCl3C(O)H while the 1,2 chlorine atom shift process accounts for the balance of the products. As expected the yield of CCl3C(O)H is oxygen dependent, increasing with the oxygen pressure. It is interesting to note that HOCHClCCl2 radicals formed by 1,2 chlorine atom rearrangement are also produced by the addition of OH to CHCl=CCl2. As discussed previously, the energised adduct radical formed by addition of OH to CHCl=CCl2 appears to eliminate HCl leading to the production of Cl2C(O) as the only stable product. However, the major reaction channel for the thermalised HOCHClCCl2O radical generated from the reaction of Cl with CCl3CH2OH appears to be carbon-carbon bond fission, and both HClC(O) and Cl2C(O) are important reaction products. The proposed reaction mechanism for the Cl atom initiated oxidation of CCl3CH2OH is shown in Figure 5. Cl + CCl3CH2OH

CCl3CHOH + HCl

O2

HO2 + CCl3C(O)H

1,2-Cl atom shift

HOCHClCCl2

O2 / X2

HOCHCl + Cl2C(O)

HOCHClCCl2O

O2

HO2 + HClC(O)

Figure 5.

Mechanism for the chlorine atom initiated oxidation of CCl3CH2OH.

Atmospheric Oxidation of the Chlorinated Solvents

179

Conclusions and atmospheric implications The major reaction pathway for HC(O)CCl2O and ClC(O)Cl2O radicals, generated from the photolysis of chloral and the Cl-atom initiated of dichloroacetyl chloride respectively is decomposition via C-C bond fission to produce phosgene. These alkoxy radicals are also proposed to be intermediates in the OH radical initiated oxidation of trichloroethene and tetrachloroethene under atmospheric conditions. Thus, the major oxidation product of 1,1,1trichloroethane, trichloroethene and tetrachloroethene is Cl2C(O). Modelling study showed that tropospheric removal of Cl2C(O) is predominantly via wet decomposition with a lifetime of around 70 days (Kindler et al., 1995). Hence Cl2C(O) may provide a small flux of Cl into the stratosphere. References Atkinson R.; Gas-phase tropospheric chemistry of organic compounds, J. Phys. Chem. Ref. Data, Monograph no.2 (1994). Kindler T.P., W.L. Chameides, P.H. Wine, D.M. Cunnold, F.N. Alyea and J.A. Franklin; The fate of atmospheric phosgene and the stratospheric chlorine loadings of its parent compounds: CCl4, C2Cl4, C2HCl3, CH3CCl3 and CHCl3, J. Geophys. Res. 100 (D1) (1995) 1235-1251. Nelson L., I. Shanahan, H.W. Sidebottom, J. Treacy and O.J. Nielsen; Kinetics and mechanism for the oxidation of 1,1,1-trichloroethane, Int. J. Chem. Kinet. 22 (1990) 577-590. Platz J., O.J. Nielsen, J. Sehested and T.J. Wallington; Atmospheric chemistry of 1,1,1-trichloroethane: UV absorption spectra and self-reaction kinetics of CCl3CH2 and CCl3CH2O2 radicals, kinetics of the reactions of the CCl3CH2O2 radicals with NO and NO2 and the fate of the alkoxy radical CCl3CH2O, J. Phys. Chem. 99 (17) (1995) 6570-6579. Sidebottom H., and J. Franklin; The atmospheric fate and impact of hydrofluorocarbons and chlorinated solvents, Pure and Appl. Chem. 68 (1996) 1757-1769. Talukdar R.K., A. Mellouki, J.B. Burkholder, M.K. Gilles, G. Le Bras and A.R. Ravishankara; Quantification of the tropospheric removal of chloral (CCl3CHO): Rate coefficient for the reaction with OH, UVabsorption and quantum yields, J. Phys. Chem. 105 (2001) 5188-5196. Tuazon E.C., R. Atkinson, S.M. Aschmann, M.A. Goodman and A.M. Winer; Atmospheric reactions of chloroethenes with the OH radical, Int. J. Chem. Kinet. 20 (1988) 241-265. Wenger J., S. Le Calvé, H.W. Sidebottom, K. Wirtz, M.M. Reviejo and J.A. Franklin; Photolysis of chloral under atmospheric conditions, Environ. Sci. Technol. 38 (2004) 831-837.

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals: Detection of Reaction Products by CW-CRDS in the Near-Infrared Range Sabine Crunaire1,3, Christa Fittschen1, Bernard Lemoine2, Alexandre Tomas3, and Patrice Coddeville3 1

PC2A, “Physico-Chimie des Processus de Combustion et de l’Atmosph è re”, UMR CNRS 8522 2 PhLAM, “Physique des Lasers, Atomes et Molécules”, UMR CNRS 8523 Université des Sciences et Technologies de Lille 1, 59655 V illeneuve d’A scq Cedex, France 3 Département Chimie-Environnement, École des Mines de Douai, 941 Rue Charles Bourseul, 59508 Douai Cedex, France Key Words: Acetic acid, CW-CRDS, OH radicals, Smog chamber

Abstract The branching ratio for the reaction of OH radicals with CH3COOD was determined at 298 r 2 K and at atmospheric pressure using an indoor smog chamber (300 L) coupled to two different detection systems: (i) Gas Chromatograph (GC) with FTIR detection for determination of [CH3COOD], (ii) Cavity Ring Down Spectroscopy (CRDS) using a telecommunication diode laser as a source for the determination of [HDO] and [H2O]. The reaction CH3COOD + OH occurs via D-atom abstraction with an efficiency of 36 r 20 %. Introduction Although anthropogenic emissions of volatile organic compounds (VOCs) in France seem to have been decreasing for several years, the contribution from oxygenated VOCs remains constant (CITEPA, 2000). Primary emissions of this type of compounds are mainly due to their use in industry (as paints, solvents, etc.). Some are also employed in fuels as substituents for aromatic hydrocarbons. In addition, as the atmospheric oxidation of hydrocarbons generally leads to oxygenated compounds (like carbonyls), secondary emissions are also significant. Among the oxygenated compounds, carboxylic acids have become the subject of growing interest over the past two decades. Low molecular weight carboxylic acids like acetic acid have been recognized as potentially important especially in urban polluted atmospheres where concentrations can exceed 20 µg/m3 (Chebbi and Carlier, 1996). Acetic acid is produced photochemically mainly from reactions of the peroxy acetyl radical (CH3CO3) with other peroxy radicals. For example, the reaction of CH3CO3 with HO2 is known to lead to about 20% CH3C(O)OH (Tyndall et al., 2001). Acetic acid in the gas phase is also produced by reaction of ozone with various olefins like propene, butene or pentene (Atkinson and Arey, 2003). A total photochemical source strength of 120 Tg/year has been reported (Baboukas et al., 2000). The contribution of direct emissions from anthropogenic (biomass combustion, motor exhaust) and biogenic (bacteria metabolisms, emission from soil and vegetation) sources is estimated at 48 Tg/year (Chebbi and Carlier, 1996). Acetic acid is principally removed from the atmosphere by wet deposition. In the atmospheric gas phase, the main loss of CH3C(O)OH is its reaction with OH radicals. 181 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 181–191. © 2006 Springer. Printed in the Netherlands.

S. Crunaire et al.

182 OH + CH3C(O)OH o products

(1)

The corresponding reaction rate, k1 = 7.5 x 10-13 cm3.molecule-1.s-1 at T = 298K (Singleton et al., 1989), corresponds to an atmospheric residence time (W k1[OH]) of more than a week (with typical tropospheric OH concentration of 1×106 molecule.cm-3). The photooxidation chain contributes to the production of photooxidants, so that the implications in the atmospheric HOx budget are important (Atkinson and Arey, 2003). However, models usually do not take these effects into account because too little is known about the fate of CH3C(O)OH in the atmosphere. In order to assess its tropospheric impact information on the product distribution of the reaction is needed. Two channels are expected to occur in the OH-initiated oxidation of acetic acid: OH + CH3C(O)OH o CH3 + CO2 + H2O o CH2C(O)OH + H2O

(1a) (1b)

According to Singleton et al. (1989) and Butkovskaya et al. (2004), hydrogen atom abstraction from the carboxyl group (pathway 1a) is the preferred pathway. A branching ratio of (64 r 17) % has been determined between 249 and 300 K (Butkovskaya et al., 2004). We present here the determination of the branching ratio of the products formed in the reaction of partly deuterated acetic acid with hydroxyl radicals under atmospheric conditions (298 K and 760 Torr) in a simulation smog chamber. OH + CH3C(O)OD o CH3 + CO2 + HDO o CH2C(O)OD + H2O

(2a) (2b)

The detection of CO2 was not possible at the time or writing due to a lack of an adequate light source (such studies are planned), however, the formation of HDO and H2O has been investigated. CH3COOH and CH3COOD have the same molecular structure, thus by analogy it is expected that the branching ratios will be fairly similar. The HDO (2a) and H2O (2b) formation rates were measuring by the CW-CRDS technique (Continuous Wave – Cavity Ring Down Spectroscopy). The setup was developed so that it could be coupled to the smog chamber. This is, to our knowledge, the first report of such a coupling and it will be described in detail in the next section. Experimental Section Simulation chamber (Figure 1) Experiments were performed in a Teflon-film bag of about 300 litres placed in a dark box equipped with an irradiation system and a temperature regulation device. The latter consist of 2 fans fixed at the top of the box flushing laboratory air around the bag. The irradiation system consists of 6 Philips TMX 204 LS fluorescent tubes (30 W, Omax = 254 nm) and 6 Philips TMX 200 LS tubes (18 W, Omax = 365 nm), distributed on both chamber sides. Each tube can be operated separately in order to modulate the irradiation flux. In the experiments presented below, only the fluorescent tubes at 365 nm have been used. Organic

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals

183

compounds are introduced into the smog chamber using a heated (T ~ 100°C) small glass chamber under vacuum. After vaporization, they are pressed into the bag by a flow of purified air (zero air). purified air

evacuation introduction of organic d

CW-CRDS GC-FTIR

heated glass chamber under vacuum

Teflon film bag

introduction of OH precursors

irradiation system: 6 fluorescent tubes at Omax = 254 nm and 6 tubes at Omax = 365 nm

Figure 1.

Top view of the indoor simulation chamber at the Ecole des Mines.

Chemicals For all the experiments, hydroxyl radicals were produced by the photolysis of methyl nitrite around 365 nm: CH3ONO + hQ (365 nm) o CH3O + NO CH3O + O2 o HO2 + HCHO HO2 + NO o NO2 + OH Methyl nitrite was prepared by slowly adding a dilute solution of H2SO4 to a mixture of NaNO2 and methanol according to the synthesis described by Taylor et al. (1980). Gas samples of about 20 mL of CH3ONO were injected in the chamber through a septum. Partly deuterated acetic acid CH3C(O)OD was obtained commercially (Acros Organics) with a purity higher than 98 % for the D-atom. Purified air was produced by flowing compressed air through a commercial air purification system (Claind AG 2301 HC). Experimental procedure Experiments were carried out at 298 r 2 K and at atmospheric pressure. Acetic acid was introduced in concentrations between 20 and 100 ppm and the bag was filled with purified air. Mixtures were left approximately one hour in the dark for stabilization during which time several samples were taken in order to determine accurately the initial concentration of reactant. The bag was then irradiated for one to four hours and sampling was

S. Crunaire et al.

184

performed at short and regular time intervals (between 15 and 30 minutes). After each run, the mixture was evacuated and the bag was then cleaned by filling it with purified air. This was repeated three times. Two different techniques were used to analyse the reaction mixture. The concentrations of CH3C(O)OD were determined using a GC-FTIR analytical device. The concentrations of reaction products HDO and H2O were measured by the CW-CRDS technique recently developed in our laboratory. Analysis apparatus 1: sample loop-TCT-GC-FTIR Sampling procedure and thermal desorption system (Figure 2) (a) Sample loop filling mode oven 150°C trap tube -190°C

(b) Trap mode oven trap tube

(c) Injection mode oven trap tube

150°C -190°C 30mL/min

50mL/min

150°C 280°C 30mL/min

sample loop

Teflon bag mixture

purified air

purified air

6 ports rotary valve

He

oven

liquid N2

He

He

empty glass tube "transfer line" cold trap tube filled with glass beads

vent

Figure 2.

GC 2 mL/min

30mL/min

GC 2 mL/min

GC 2 mL/min

Schematic diagrams of the device for sampling and thermal desorption.

Gas samples were taken from the bag using a 20 mL loop (4.57 mm i.d. Silcosteel tubing) and a 6-port valve (Valco). Figure 2a shows the loading position of the valve. The gas sample from the chamber is pumped through the heated (T = 100°C) sample-loop at 50 mL/min for 4 min. After switching off the pump, the 6-port valve is turned to the injection position (Figure 2b) and the 20 mL gas sample is conveyed to the cold trap of the TCT Chrompack (Thermal desorption and Cold Trap) device by the carrier gas. It should be noted

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals

185

that this system allows several fillings of the sample loop to increase the sampling volume. During the experiments, depending on the concentrations of reactants, injections of one or two loops have been carried out. The cryo-focusing with liquid nitrogen was enhanced for the more volatile compounds by filling the trap (uncoated deactivated fused silica; i.d. = 0.53 mm; length = 40 cm) with glass beads. Finally, Figure 2c shows the injection step. The trap tube was subject to a fast temperature increase (approximately 10°C/s) in order to inject the compounds into the gas chromatograph within a few seconds. GC-FTIR analysis The sample loop-TCT injection unit is coupled to a gas chromatograph (GC) Varian 3300 with detection by a FTIR spectrometer (Nicolet Magna 550) (Figure 3). The chromatographic separation was performed using a capillary column CP SIL 5 CB (50 m x 0.32 mm i.d.). The following temperature program was used for the GC oven: isothermal at 40°C for 5 min, first temperature ramped at 8°C/min to 120°C, second temperature ramped at 15°C/min to 225°C, where it was maintained for 5 min. Helium was used as the carrier gas at an initial pressure of 15 psi (40°C). The column effluent was passed through a fused silica transfer line (0.6m x 0.32mm i.d.) to the light pipe, both maintained at 240°C. The light pipe is a narrow borosilicate tube with a smooth, thin gold coating layer on the inside surface. By reflecting the IR beam through the light pipe via this coating, the path length of the cell is increased by a factor of ten or more with respect to the actual length of the tube (Ragunathan et al., 1999). The infrared beam was focused onto a fast response MCT detector cooled with liquid nitrogen.

TCT

Gas sample loop (20mL – T ~100°C)

Michelson interferometer

FID

capillary column

Teflon bag's sample

lightpipe

IR source

Principal infrared bench

Figure 3.

MCT detector

GC-FTIR interface

Transfer line

TCT-GC-FTIR analytical device.

GC

S. Crunaire et al.

186

Data acquisition was performed using compatible Omnic software (Nicolet). The Gram-Schmitt (total spectral response) chromatogram was constructed from interferograms registered at different times. Spectra were taken in real time between 650 and 4000 cm-1 at a resolution of 16 cm-1 with a scan every 2 s (average of 10 interferograms). Functional group chromatograms were extracted from the Gram-Schmitt chromatogram in the 1600-1900 cm-1 spectral region (C=O band) to carry out the quantitative analysis of acetic acid. Analysis apparatus 2: CW-CRDS Introduction Cavity Ring-Down Spectroscopy was introduced in 1988 by O'Keefe and Deacon as a spectroscopic method for absorption measurements (O'Keefe and Deacon, 1988). It is a versatile high sensitivity absorption technique. One of the most essential advantages of CRDS in contrast to usual absorption methods is that the CRDS signal is not affected by intensity fluctuations of the laser since only the decay time of the signal, which does not depend on the laser intensity, is detected. The setup of O'Keefe and Deacon employed a pulsed, tunable laser, and is known as a "generic pulsed CRDS setup". Since the publication of their article, the method of CRDS has been applied to different absorptions ranging from ~ 200 nm to ~ 10 µm and many variants of excitation and detection schemes have been used. Various methods and applications of CRDS are described in a number of reviews (Wheeler et al., 1998; Berden et al., 2000; Brown, 2003). In 1997, Romanini et al. adapted the use of small, continuous-wave diode lasers to CRDS (Romanini et al., 1997a; Romanini et al., 1997b). In order to perform quantitative analyses of H2O and HDO during oxidation of deuterated acetic acid in the smog chamber, we have developed a similar setup. CRDS principle The principle of CRDS is based upon a measurement of the rate of decay of an optical resonator with a high quality factor, indicating a long lifetime of the photons in the cavity. The principle is schematically illustrated in Figure 4. L

L

laser

laser

.. .

.. . M1

M2

(a) Figure 4.

time

M1

d

(b)

M2

time

Principle of CRD absorption measurements. (a) light decaying in an empty cavity, (b) cell is filled with an absorber.

After coupling the laser light into the cavity, the light intensity decreases with every pass within the resonator because of reflection losses of the mirrors and absorption losses of the medium. Therefore an exponential decay is obtained at the detector (eq.1).

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals § t· It I 0 ˜ exp¨  ¸ © W¹

187

(eq.1)

Without absorbing species between the mirrors (Figure 4a), the ring down time of the optically empty cavity is given by (eq.2): W0

L c ˜ (1  R )

(eq.2)

where c is the velocity of light and R the reflectivity of the mirrors. The decay time of the cavity filled with an absorbing medium (Figure 4b) is (eq.3): W abs

L c ˜ >1  R  Dd @

(eq.3)

where Dthe linear absorption coefficient is inversely proportional to the measured decay time of the CRDS signal (eq.4; here L=d).

D

1 § 1 1· ˜ ¨¨  ¸¸ c © W abs W 0 ¹

(eq.4)

Using D makes it possible to determine directly the concentration of a trace gas inside the cavity if the absorption cross-section Vabs is known (eq.5).

D

>abs@˜ V abs

(eq.5)

CW-CRDS setup and data acquisition

The experimental setup (Figure 5) used in this work was very similar to that described by Romanini et al. (1997b). It works with an external cavity diode laser (Sacher Lasertechnik) tunable from 1420 to 1480 nm with an output power of at least 3 mW. For the cavity mirrors, a 1 m radius of curvature and low-loss mirrors by Layertec, covering the range from about 1400 nm to 1550 nm, were used. They were placed at a distance L of 70 cm and sealed the cell. One of them was mounted on a piezoelectric transducer (Physik Instrumente). This device was used to modulate the cavity length in order to let one of the cavity modes oscillate around the laser line. As the beam reflected by the input cavity mirror can strongly perturb the laser, a Faraday isolator (Isowave) was placed in the path to attenuate this effect. The intensity of the light leaving the cavity was measured with a photodiode avalanche. When a predefined threshold was reached, an acoustic-optical modulator (Gooch and Housego) turned off the laser beam and a clean ring-down decay could be measured. The ring-down signals were registered by an InGaAs photodiode (EG&G) and digitised by a 12 bits PC interface card (National Instruments). Using the graphic programming language LabVIEW (National Instruments), the ring-down signals were recorded at a rate of about 100 Hz. The analysis of each ring-down (fitted by an exponential function using a nonlinear Levenberg-Marquardt algorithm) is made separately before averaging the decay time W. With the system described above, a section of the C2H2 overtone spectrum around 1512 nm has been measured, using a distributed feedback diode laser. The noise equivalent absorption coefficient is in the order of 5×10-8 cm-1 or a detection limit of less than 100 ppb

188

S. Crunaire et al.

for C2H2 under the experimental conditions. A summary of the performance characteristics of the setup is given in Table 1.

1 meter

Wavemeter

1468.584 nm

0.75 meter

Fiber optic optic Fiber L

ECDL

AOS

OI

Trigger system Gate generator

1st order

L

L

L PZT

Glass cell APD T, P L

MC

M CC

P

PC

N2 or He

Threshold detector

gas or or smog smog Standard gas chamber's sample sample chamber's

A/D Converter working with LabVIEW 6.1TM

ECDL: External Cavity Diode Laser; OI: Optical Isolator; L: Lens; AOS: Acousto-Optic Switch; PZT: Piezo-Electric actuator; MC: Cavity mirror; APD: Avalanche Photodiode; PC: Pressure Controller; P: Pump

Figure 5.

Scheme of the CW-CRDS setup

Table 1.

Summary of the performance characteristics of the CW-CRDS setup.

coarse tunability of the ECDL fine tunability of the ECDL cavity length cavity mirrors

between 1420 and 1490 nm on 1 nm using the grating piezo control input 70 cm with actual cell RC = 1 m; R ~ 99.99 % o W0 = 25 µs

~ 7 km without absorbing species frequency of cavity length > 200 Hz o one ringmodulation down event every 2.5 ms spectral resolution > 0.01 cm-1 effective path length

sensitivity of the system

5.10-8 cm-1

Determination of [HDO] and [H2O] during smog chamber experiments For quantitative determination of the HDO and H2O concentrations before and during oxidation of CH3COOD by OH in the smog chamber, the reaction mixture was pumped with a flow rate of 70 mL/min through the cavity. During a measurement, the cavity pressure is held constant at 50 Torr to reduce the broadening of the two spectral lines and thus increase the selectivity. Figure 6 shows a CW-CRDS spectrum of characteristic spectral lines of H2O and HDO near 6776 cm-1 (without any averaging). From this measurement, the corresponding concentrations are calculated using the database HITRAN 2004 (Rothman et al., 2005). It takes approximately one minute to measure a spectrum as presented in Figure 6, however, for real-time detection it is sufficient to measure the absorption at only a few wavelengths. Thus

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals:

189

the concentrations can be determined within a few seconds. Note that quantitative analysis does not require any calibration, CW-CRDS leads directly to absolute concentrations, if the absorption cross section of the measured spectral line is known (from HITRAN 2004 or GEISA databases for example).

+ experimental points Gaussian fit

HDO

H2 O

Figure 6.

CW-CRDS spectrum obtained at 50 Torr from a smog chamber mixture.

Preliminary Results and Discussion The study of the reaction OH + CH3COOD was conducted in an excess of CH3ONO over CH3COOD. The branching ratio of channel (2a) was determined as the ratio of the concentration of HDO produced to the concentration of CH3COOD consumed in reaction 2: R

k 2a k 2a  k 2b

k 2a k2

d [ HDO ]  d [ CH 3 COOD ]

>HDO @formed [ CH 3 COOD ] reacted

Figure 7 shows an example of a plot of the formation of HDO and H2O versus the loss of CH3COOD for one experiment. The accuracy of the measured CH3COOD concentration was determined from the statistic error (2V) of the calibration curve, obtained by liquid injections. It was estimated to be 5 %. Uncertainties in the HDO and H2O concentrations arise mainly from errors in the measured absorbance and from the accuracy in the absorptions cross sections. They were estimated to be 10 %. The slopes of the above lines give directly the branching ratios R2a and R2b. At first, the branching fraction R2a = (36 r 20) % indicates that channel 2a would be the minor oxidation channel for the OH + CH3COOD reaction. However, this result can't be easily compared with the one of Butkovskaya et al. (2004) because too little is known about the

190

S. Crunaire et al.

isotope effect in the reaction of CH3COOD and OH. On the one hand, the yield of formation of H2O is too high (R2b > 1) and other sources of production must be taken into account. Principally, the interactions between H2O molecules and the walls of the chamber can be responsible for a large perturbation in the H2O absorption signal. Furthermore, the reactions of OH with other species like CH3ONO, HCHO (product of CH3ONO photolysis) or CH3OH (impurity in the CH3ONO synthesized) can be supplementary sources of H2O.

3 '[HDO] and '[H2O] (molecules/cm )

4,5E+15 4E+15

HDO

3,5E+15

H2O

y = 2,2763x R2 = 0,983

3E+15 2,5E+15 2E+15 1,5E+15

y = 0,3627x

1E+15

R2 = 0,6895

5E+14 0 0

2E+14

4E+14

6E+14

8E+14

1E+15

1,2E+15 1,4E+15 1,6E+15 1,8E+15

-'[CH3COOH] (molecules/cm3)

Figure 7.

Formation of HDO (crosses) and H2O (open circles) versus loss of CH3COOD following irradiation of a CH3COOD (~ 3,5×1015 molecules/cm3) + CH3ONO (~ 4×1018 molecules/cm3) mixture with zero air as diluent at 298 r 2 K and atmospheric pressure. The solid lines are linear regressions of the two data sets.

Conclusions The present study reports the first coupling between a smog chamber and a CW-CRDS setup. After a successful evaluation the CW-CRDS setup in the laboratory (spectroscopic study of C2H2 around 1512 nm), this new analytical device permits us to follow the production of small molecules (CO, CO2, C2H2, HCOOH, etc.) and also of isotopic compounds (HDO, H218O for example) almost in "real time" during experiments on VOC degradation (oxidation or photolysis) in a simulation chamber. For the first experiments carried out on the oxidation of CH3COOD by OH radicals, a branching ratio of (36 r 20) % has been determined for the D-atom abstraction channel. These experiments have revealed several difficulties in the analysis of the results and comparison with literature data because of the unknown reaction mechanisms and kinetics involving deuterated compounds. Additional experiments are necessary to resolve these difficulties. Determination of the branching ratio for the product channels in reaction of CH3COOH with OH radicals in a smog chamber under atmospheric conditions by measuring the CO2 formation yield (CW-CRDS using a distributed feedback diode laser at 1570 nm) is in progress.

Simulation Chamber Study of the Oxidation of Acetic Acid by OH Radicals:

191

Acknowledgments The authors gratefully acknowledge the help of T. Leonardis (Ecole des Mines – Douai) and Dr. D. Romanini (Laboratoire de Spectrométrie Physique – Grenoble) for their technical assistance. Moreover, this work has been achieved with the supports of the CPER (Contrat de Plan Etat Région) Nord-Pas de Calais "air quality", the CNRS (Centre National de la Recherche Scientifique), the CERLA (Centre d'Etudes et de Recherches Lasers et Applications) and the FEDER (Fonds Européen de DEveloppement Régional). References Atkinson R. and J. Arey; Atmospheric degradation of volatile organic compounds, Chem. Rev. 103 (2003) 46054638. Baboukas E. D., M. Kanakidou and N. Mihalopoulos; Carboxylic acids in gas and particulate phase above the Atlantic Ocean, J. Geophys. Res. 105 (2000) 14459-14471. Berden G., R. Peeters and G. Meijer; Cavity ring-down spectroscopy: Experimental schemes and applications, Int. Rev. in Phys. Chem. 19 (2000) 565-607. Brown S. S.; Absorption spectroscopy in high-finesse cavities for atmospheric studies, Chem. Rev. 103 (2003) 5219-5238. Butkovskaya N. I., A. Kukui and N. Pouvesle, G. Le Bras; Rate constant and mechanism of the reaction of OH radicals with acetic acid in the temperature range of 229-300 K, J. Phys. Chem. A 108 (2004) 7021-7026. Chebbi A. and P. Carlier; Carboxylic acids in the troposphere, occurrence, sources, and sinks: a review, Atmos. Environ. 30 (1996) 4233-4249. CITEPA : Emissions dans l’air en France, comparaisons de données d'émissions de la France et d'autres pays, URI: 'http://www.citepa.org/emissions/france_autres/Comparaisons.pdf' (2000). HITRAN 2004; URI: 'http://www.hitran.com'. O'Keefe A. and D. A. G. Deacon; Cavity ring-down optical spectrometer for absorption measurements using pulsed laser sources, Rev. Sci. Instrum. 59 (1988) 5244-2551. Ragunathan N., K. A. Krock, C. Klawun, T. A. Sasaki and C. L. Wilkins; Gas chromatography with spectroscopic detectors, J. Chromatogr. A 856 (1999) 349-397. Romanini D., A. A. Kachanov, N. Sadeghi and F. Stoeckel; CW cavity ring down spectroscopy, Chem. Phys. Lett. 264 (1997a) 316-322. Romanini D., A. A. Kachanov and F. Stoeckel; Diode laser cavity ring down spectroscopy, Chem. Phys. Lett. 270 (1997b) 538-545. Rothman L.S., D. Jacquemart, A. Barbe, D. Chris Benner, M. Birk, L.R. Brown, M.R. Carleer, C. Chackerian Jr., K. Chance, V. Dana, V.M. Devi, J.-M. Flaud, R.R. Gamache, A. Goldman, J.-M. Hartmann, K.W. Jucks, A.G. Maki, J.-Y. Mandin, S.T. Massie, J. Orphal, A. Perrin, C.P. Rinsland, M.A.H. Smith, J. Tennyson, R.N. Tolchenov, R.A. Toth, J. Vander Auwera, P. Varanasi, and G. Wagner; The HITRAN 2004 Molecular Spectroscopic Database, JQSRT, Preprint version, 2005. Singleton D. L., G. Paraskevopoulos and R. S. Irwin; Rates and mechanism of the reactions of hydroxyl radicals with acetic, deuterated acetic, and propionic acids in the gas phase, J. Am. Chem. Soc. 111 (1989) 52485251. Taylor W. D., T. D. Allston, M. J. Moscato, G. B. Fazekas, R. Kozlowski and G. A. Takacs; Atmospheric photodissociation lifetimes for nitromethane, methyl nitrite and methyl nitrate, Int. J. Chem. Kinet. 12 (1980) 231-240. Tyndall G. S., R. A. Cox, C. Granier, R. Lescaux, G. K. Moortgat, M. J. Pilling, A. R. Ravishankara and T. J. Wallington; Atmospheric chemistry of small organic peroxy radicals, J. Geophys. Research. 106 (2001) 12157-12182. Wheeler M. D., S. M. Newman, A. J. Orr-Ewing and M. N. R. Ashfold; Cavity ring-down spectroscopy, J. Chem. Soc., Faraday Trans. 94 (1998) 337-351.

Kinetics, Products and Mechanism of O(3P) Atom Reactions with Alkyl Iodides Ian Barnes Bergische Universität Wuppertal, Fachbereich C - Physikalische Chemie, Gauß Straße 20, D-42119 Wuppertal, Germany

Key Words: Alkyl iodides, Halogens, Oxygen atoms, Relative kinetic, Reaction mechanisms

Introduction Alkyl halides (RX: X = Cl, I) are an important source of halogens in the atmosphere. The major tropospheric sinks of these compounds are photolysis (RBr, RI) and reaction with OH radicals. In the case of alkyl iodides (RI) relative kinetic studies of their OH reactions in photoreactors are complicated by fast reactions with the O(3P) atoms generated by the photochemical OH radical sources. Figure 1 below shows a ln-ln plot of the kinetic data from an experiment performed in a large photoreactor to determine the OH rate coefficient for the reaction OH + CH3CH2CH2I relative to OH + ethene using the photolysis of methyl nitrite (CH3ONO) as the OH radical source. A recent example of the implementation of the relative kinetic technique for the determination of OH radical rate coefficients in a photoreactor can be found in Olariu et al. (2000).

ln((Io/I) propyl iodide

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.00

0.10

0.20

0.30

0.40

0.50

0.60

ln(Io/I) ethene

Figure 1.

Plot of the kinetic data obtained from the irradiation of a propyliodide, ethene, CH3ONO, and synthetic air reaction mixture in a 420 l photoreactor.

The non-linearity with increasing reaction time in the plot in Figure 1 is due to very fast reactions of RI with O(3P). An analysis of the initial slope yields a rate coefficient for the reaction of OH with 1-propyliodide which is in reasonable agreement with the recommended literature value for the reaction of OH with 1-propyliodide (Carl and Crowley, 2001; Cotter et al., 2003) while the slope at extended reaction time yields a rate coefficient in reasonable agreement with literature values for the reaction of O(3P) with 1-propyliodide (Gilles et al., 1996; Terual et al., 2004). Luo et al. (1995) have successfully employed the relative kinetic technique in a 500 l Teflon chamber to measure rate coefficients for the reactions of O(3P) with a series of alkenes 193 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 193–205. © 2006 Springer. Printed in the Netherlands.

194

I. Barnes

and terpenes at ambient temperature and atmospheric pressure. The experiments were performed in N2 bath gas and an excess of both NO2 and NO to optimise the concentration of O(3P) and depress possible interferences by reactions of OH and NO3 radicals and O3 with the compounds under investigation. The few rate coefficients for the reactions of O(3P) reactions with alkyl iodides, which have been measured, have been found to be extremely rapid (Gilles et al., 1996; Terual et al., 2004; and references therein). The measured rate coefficients are much faster than those for O(3P) with the chlorine and bromine analogs and are also much faster than those for OH radicals with the alkyl iodides (Carl and Crowley, 2001; Cotter et al., 2001). This has been explained by invoking the initial formation of a R–I–O complex. IO elimination is the most important product channel for the small iodoalkanes such as CH3I and CF3I, however, for longer chain iodoalkanes the presence of a C–H bond in the E-position to the I-atom enables the intramolecular abstraction of an H-atom via a five-membered ring transition state to eliminate HOI and form an alkene (Klaassen et al., 1996; Loomis et al., 1997). As discussed above the kinetic database for the reactions of O(3P) with alkyl iodides is relatively limited. The higher reactivity of alkyl iodides to O(3P) atoms compared to OH radicals suggests that the application of the relative kinetic technique in a photoreactor, as applied by Luo et al. (1996) for the determination of O + alkene/terpene rate coefficients, should also be applicable for the determination of rate coefficients for O + alkyl iodides. Because of the much higher rate coefficients for O(3P) with the alkyl iodides compared to OH, inference by any OH radicals produced in the system will be minimal. Reactions with O3 and NO3 radicals are too slow to be of any importance. In the present work the relative kinetic technique has been applied in large and small photoreactors to measure rate coefficients for the reaction of O(3P) atoms with a series of alkyl iodides at room temperature and atmospheric pressure. The products formed in N2 have also been investigated. Experimental Product and kinetic experiments have been performed on the reaction of O(3P) with a series of mono- and disubstituted alkyl halides. The compounds investigated are listed in Table 1 in the Results section. The product studies were all carried out in either a 1080 1 quartz glass chamber (Mihalopoulos et al., 1992) or a 480 l cylindrical Duran glass chamber (Barnes et al., 1993). Schematic outlines of the 1080 l
and 480 l chambers are shown in Figures 2 and 3, respectively. In the 1080 l
chamber an internally mounted White type multiple reflection mirror system operated at a total optical path length of (484.7 r 0.8) m coupled to an FTIR spectrometer was used for reactant and reference monitoring. IR spectra were recorded at a spectral resolution of 1 cm-1 using a Nicolet NEXUS FT-IR spectrometer equipped with a MCT detector. The 480 l chamber is also equipped with 3 built-in White mirror systems for long path (51.6 m) FTIR (Nicolet Magna 520, MCT detector, 1 cm-1 resolution), VIS (28.8 m) and UV (28.0 m) in situ monitoring of reactants and products. The majority of the kinetic experiments have been carried out in a simple 20 l reactor using GC-FID for the analysis; Hewlett Packard 5890 Series II with HP 3395 integrator, separation on a RESTEX RTX-1301 column (30 m long, 0.53 mm inner diameter and 0.25 Pm film thickness). This reactor is shown in Figure 4. Samples were taken using a 1 ml gas-

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

195

tight syringe and injected into the GC system. For some of the larger alkyl iodides control experiments were also performed in the 480 l chamber using FT-IR absorption spectroscopy for the kinetic analysis.

Figure 2.

Schematic outline of a 1080 l quartz glass photoreactor equipped with an internally mounted White type mirror system for FTIR in situ monitoring of reactants and products. The chamber can be temperature regulated in the range 303 to 283 K. The chamber is surrounded by 32 super actinic fluorescent lamps (Philips TL 05/40 W: 320 < O < 480, Omax = 360 nm) and 32 low pressure mercury lamps (Philips TUV 40 W: Omax = 254 nm) mounted evenly around the reactor. Other types of lamps can be easily mounted.

All experiments were performed at 1013 mbar total pressure of N2 at 298 ± 2 K using the photolysis of NO2 with fluorescent lamps (Philips TL 40W/05, 320 < O < 450 nm) as the O(3P) atom source: NO2 + hv • NO + O Rate coefficients for the reaction of O(3P) with selected alkyl iodides were determined using the relative kinetic technique as outlined by Luo et al. (1996). 2-Iodopropane (CH3CH(I)CH3) was used as the reference compound: O + RI o products; k1 O + 2-iodopropane o products; k2

I. Barnes

196

Providing that RI and 2-iodopropane are removed solely by reaction with O atoms the following rate law holds: ­[ RI]t 0 ½ ln ® ¾ ¯ [RI]t ¿

k1 ­[2 - iodopropane ]t 0 ½ ln ® ¾ k 2 ¯ [2 - iodopropane ]t ¿

(eq 1)

where [RI]to and [2-iodopropane]to are the concentrations of the alkyl iodide under test and reference alkyl iodide, respectively, at time to, [RI]t and [2-iodopropane]t are the corresponding concentrations at time t, and k1 and k2 are the rate coefficients of the alkyl iodide and reference compound with O(3P), respectively. Plots of ln([RI]to/[RI]t) against ln([2iodopropane]to/[2-iodopropane]t) should straight lines with a slope of k1/k2 and zero intercept. The rate coefficients for the reactions of O(3P) with the various alkyl iodides were placed on an absolute basis using a rate coefficient value of k = 4.97 × 10-11 cm3 molecule-1 s-1 from Teruel et al. (2004) for the reaction of O(3P) with 2-iodopropane. This value is in good agreement with the other reported value by Gilles et al. (1996).

Figure 3.

Schematic outline of a 480 Aquartz glass photoreactor equipped with an internally mounted White type mirror systems for FTIR, VIS and UV in situ monitoring of reactants and products. The chamber is surrounded by 20 super actinic fluorescent lamps (Philips TL 05/40 W: 320 < O < 480, Omax = 360 nm) and 12 gold fluorescent lamps (Philips TLD 16/36 W: 500 < O < 700 nm).

The approximate initial reactant concentrations in both reactors (in ppmv units) were as follows: NO2, 50-100; alkyl iodide 10-30 and 2-iodopropane 10-30 (1 ppmV = 2.46 u 1014 molecule cm-3 at 298 K and atmospheric pressure). The concentration-time behaviour of the alkyl iodides were monitored over a 15-30 min irradiation period in both the kinetic and product studies.

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

197

Figure 4. Schematic outline of a 20 l glass reactor with a metal top flange for inlet and outlet ports. The reactor is equipped with 4 fluorescent lamps (Philips TL05/40 W) and can be evaluated to 0.01 mbar with a simple rotary pump.

Figure 5. Photograph of the simple 20 l glass reactor with GC analysis system shown schematically in figure 4.

Figure 5. Photograph of the simple 20 l glass reactor with a GC analysis system shown schematically in Figure 4.

Results and Discussion Kinetics O + RI To test the relative kinetic method in the small 20 l reactor the rate constant ratio for the reaction of O(3P) with 1-iodopropane relative to 2-iodopropane was determined. The kinetic data from a typical experiment plotted according to equation (1) are shown in Figure 6. A good linear correlation with zero intercept was obtained. The value of the rate coefficient obtained for the reaction of O(3P) with 1-iodopropane is in excellent agreement with the absolute value obtained by Teruel et al. (2004) using pulsed laser photolysis-resonance Fluorescence (PLP-RF) which demonstrates the validity of the applied method for the determination of rate coefficients for the reactions of O(3P) with alkyl iodides.

I. Barnes

198 0.70

Figure 6. Plot of the kinetic data according to eq. (1) for measurements on the reaction of O(3P) with 1iodopropane relative to that with 2-iodopropane.

ln(Io/I) 1-iodopropane

0.60 0.50 0.40 0.30 0.20 0.10 0.00 0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

ln(Io/I) 2-iodopropane

Examples of the kinetic data for some of the alkyl iodides investigated plotted according to eq. (1) are shown in Figure 7. In the main good straight line correlations were obtained with zero intercepts. The rate coefficients obtained for the alkyl iodides investigated are listed in Table 1 where they are compared with the available literature data. The rate coefficients determined in this work are estimated to have a total error of ± 15% 3.50

ln(Io/I) alkyl iodide

3.00 2.50 2.00

iodocyclopentane 1,4-diiodobutane 2-iodo-2-methylpropane 1-iodo-2-methylpropane 1,3-diiodopropane iodocyclohexane Linear (1,4-diiodobutane) Linear (iodocyclopentane) Linear (2-iodo-2-methylpropane) Linear (1-iodo-2-methylpropane) Linear (iodocyclohexane) Linear (1,3-diiodopropane)

Figure 7. Examples of plots of the kinetic data according to eq (1) for measurements on the reaction of O(3P) with several alkyl iodides relative to that with 2iodopropane.

1.50 1.00 0.50 0.00 0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

ln(Io/) 2-iodopropane

Rate coefficients have been determined for the reaction of O(3P) with 17 alkyl iodides 12 of which are first-time measurements. As can be seen from Table 1 there is good agreement between the rate coefficients reported in the literature for the C1 to C4 alkyl monoiodides (Gilles et al., 1996; Teruel et al., 2004) and those determined in this work. This good agreement adds further weight to the soundness of the applied experimental approach. The present work has considerably expanded the number and type of alkyl iodides investigated. The reactivity trends for the reactions of O(3P) with RI have been discussed in detail by Teruel et al. (2004) and Gilles et al. (1996). Teruel et al. (2004) have also made comparisons with the analogous reactions with OH radicals and Cl atoms. Therefore, only salient features of the reactivity trends and supplementary information will be presented here.

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

199

As in the other studies a gradual increase in the rate coefficient is observed in progressing from CH3I to CH3CH2I to CH3CH2CH2I. It has been observed in this study that introducing a fourth carbon to the linear chain results in a further small increase in the rate coefficient, however, extensions of the chain to C5 and C6 have no further influence on the rate coefficient. In agreement with Teruel et al. (2004) it is found that the rate coefficient for the reaction of O(3P) with 2-iodopropane is greater than that for 1-iodopropane. It has also been found in the present work that the rate coefficient for the reaction of O(3P) with 2-iodobutane is higher than that for 1-butane. This behaviour is reconcilable in terms of the proposed main reaction mechanism mentioned in the Introduction, i.e. the secondary iodides have a larger number of E H-atoms that can be internally abstracted via a five-membered ring R-I-O complex transition state. For the C5 compounds 1-iodo-2-methylbutane and 1-iodopentane the rate coefficient for O(3P) with 1-iodo-2-methylbutane is higher than that for 1-iodopentane, presumably for similar reasons. The highest rate coefficient has been found for the reaction of O(3P) with 2-iodo-2-methylpropane which has the highest possible number abstractable E Hatoms. A number of di-iodo compounds have been investigated. In general the rate coefficients for the di-iodo compounds with chain lengths of •C2 are about a factor of 2 higher than the corresponding mono-iodo compounds. In the case of CH3I and CH2I2 the difference reported by Teruel et al. (2004) for the rate coefficients is nearly 4. The rate coefficient for the reaction of O(3P) with neopentyl iodide ((CH3)3CCH2I) is slightly higher than that for CH3I. Neopentyl iodide does not possess any E H-atoms and the mechanism is, as in the case of CH3I, an overall I abstraction mechanism to form IO. This conclusion is supported by the observations in the product studies. The rate coefficients for the reaction of cyclopentane and cyclohexane with O(3P) are larger than those for the corresponding linear chain 1-iodo-alkanes, which probably reflects the lower C-H bond strength in the cyclic compounds compared to the straight chain analogues. The measured rate coefficient for the reaction of O(3P) with cyclopentane in this work is, however, around 30% higher than that for cyclohexane despite the C-H bond energy in cyclohexane (399.6 kJ mol-1) being somewhat lower than that in cyclopentane (403.5 kJ mol-1). Stereo factors may possibly play a role in determining the rate of the reaction; cyclohexane has a chair conformation with zero ring strain whereas cyclopentane has an envelope conformation. This is, however, presently only conjecture, independent conformation of the rate coefficients for the two substances by another method is required before any positive conclusions can be drawn. The room temperature rate coefficients for O(3P) with fluorinated alkyl iodides (Gilles et al., 1996) and selected alkyl iodides (Teruel et al., 2004) and have found to show a linear correlation with the ionisation potential (IP) of the compounds. Table 2 lists the ionization potentials (IP) for a series of alkyl iodides taken from the NIST Chemistry WebBook (http://webbook.nist.gov/chemistry/) and their corresponding rate coefficients for reaction with O(3P) as measured in this study and taken from other available literature sources. The natural logarithm of the 298 K rate coefficients for– the compounds listed in Table 2 are plotted as a function of ionisation potential in Figure 8. The numbers on the plot refer to the correspondingly numbered compounds in Table 2. The points for compounds 3 and 9 overlap on the diagram.

I. Barnes

200 Table 1.

Summary of the product and kinetic data from studies on the reactions of O(3P) with alkyl halides at 1000 mbar total pressure of N2 and 298 ± 2K. The rate coefficients determined in this work (column (a)) have an estimated overall uncertainty of ± 15%. The alkene yields have an uncertainty of approximately ± 20%. k(O + RI) x 1011

Compound

Products

(cm3 molecule-1 s-1)

(a)

(b)

(c)

Alkenes (d) / (% mol yield)

Alkanes/other

iodomethane

CH3I

1.82 2.01

-

-

ethane

iodoethane

CH3CH2I

3.19 3.51

-

ethene / 100

-

1,2-diiodoethane

ICH2CH2I

6.25

-

Iodoethane / (f)

?

-

ethene / 42 1-iodopropane

CH3CH2CH2I

3.88 3.79

-

2-iodopropane

(CH3)2CHI

ref

1,3-diodopropane

ICH2CH2CH2I

11.5

-

-

allyliodide / 100

allyliodide

CH2=CHCH2I

10.2

-

-

1,5-hexadiene / (f)

3-iodopropylene oxide

1-iodobutane

CH3CH2CH2CH2I

4.83

-

-

1-butene / 70

n-octane

2-iodobutane

CH3CH2CHICH3

6.64

-

-

1-butene / 46

3,4-dimethylhexane

4.97 5.18

propene / 80

n-hexane

propene / 68

2,3-dimethylbutane

trans-2-butene / <10

1-iodo-2methylpropane

(CH3)2CHCH2I

4.39

-

4.54

isobutene / 65

2,5-dimethylhexane

2-iodo-2methylpropane

(CH3)2C(I)CH3

14.3

-

-

isobutene / 65

2,2,3,3-tetramethylbutane

8.42

-

-

4-iodo-1-butene / (f)

4-iodo-1,2-butylene oxide

1,4-diiodobutane ICH2CH2CH2CH2I

butadiene / 55(e)

1-iodopentane

CH3(CH2)4I

4.50

1-iodo-2methybutane

CH3CH2CH(CH3)CH2I 5.36

neopentyliodide

(CH3)3CCH2I

-

-

1-pentene / (f)

n-decane

2-methyl-1-butene / 100

3,6-dimethyloctane

2.50

-

-

(g)

2,2,5,5-tetramethylhexane

iodocyclopentane cyclo-C5H9I

9.27

-

-

cyclopentene/ (f)

bicyclopentyl

1-iodohexane

CH3(CH2)5I

4.75

-

-

1-hexene/ (f)

n-dodecane

iodocyclohexane

cyclo-C6H11I

6.91

-

-

cyclohexene / 68%

(a) This work; the values were put on an absolute basis using k = 4.97 x 10-11 cm3 molecule-1 s-1 for O(3P) with 2-iodopropane (ref) from Teruel et al. (2004); (b) Values from Teruel et al. (2004); (c) Values from Gilles et al. (1996); (d) Yield corrected for secondary reaction with O atoms; (e) Yield increased with time; the given yield is the yield at end of the experiment; (f) Observed but yield not determined; (g) No evidence for the formation of an alkene.

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

201

The points in Figure 8 taken collectively indicate a distinct curvature in the relationship. However, if the points for the fluorinated compounds 14 and 16 are removed a reasonably linear correlation between the points for the remaining compounds is obtained.

-1 -1

rate coefficient (cm molecule s )

1.00E-09

4

1.00E-10 6

7

3

2

5 10 9 3

8 12 11

1.00E-11

13

14

16

15

1.00E-12 8.8

9

9.2

9.4

9.6

9.8

10

10.2

10.4

10.6

10.8

ionisation potential (eV)

Figure 8.

Plot of the rate coefficients for the reactions of O(3P) with a series of alkyl iodides as a function of the ionisation potential of the alkyl iodide. The numbers on the plot refer to the correspondingly numbered compounds in Table 2.

Products O + RI With the exception of iodomethane (CH3I) and neopentyl iodide ((CH3)3CCH2I) all the reactions of O(3P) with the alkyl iodides were found to produce as major product an alkene. An example is shown for 1-iodobutane in Figure 9. Hypoiodous acid (HOI), the probable coproduct in each case, was also identified via its known infrared absorptions at 3620 and 1070 cm-1 (Barnes et al., 1992). This compound is, however, very short-lived under the conditions of the experiments and was only visible in the infrared spectrum in the initial stages of the reaction (Figure 10) and could therefore not be quantified. For many of the alkenes calibrated reference spectra were available and the yield of the alkene has been determined. Oxygen atoms also react fairly rapidly with alkenes, therefore, the reaction of O atoms with the alkenes has been taken into consideration when calculating the alkene yields from the O + RI reactions. The yields and the identity of the alkene for the respective O + RI reactions are listed in Table 1.

I. Barnes

202 Table 2.

Comparison of the ionization potentials (IP) of selected alkyl iodides with the measured rate coefficients for the compounds with O(3P) atoms. The IPs’ are taken from the NIST Chemistry WebBook (http://webbook.nist.gov/chemistry/).

No Compound

Formula

1

neopentyliodide

(CH3)3CCH2I

2

1-iodohexane

CH3CH2CH2CH2CH2CH2I

3

1-iodopentane

4

Ionization Potential (eV)

rate coefficient k(O3P) 3 (cm molecule-1 s-1)

not known

2.60 · 10-11

9.179

4.75 · 10-11

CH3CH2CH2CH2CH2I

9.2

4.50 · 10-11

2-iodo-2-methylpropane

(CH3)2C(I)CH3

8.98

1.49 · 10-10

5

1-iodobutane

CH3CH2CH2CH2I

9.23

5.06 · 10-11

6

2-iodobutane

CH3CH2CHICH

9.1

6.84 · 10-11

7

2-iodopropane

(CH3)2CHI

9.175

4.97 · 10-11

8

1-iodopropane

CH3CH2CH2I

9.26

3.79 · 10-11

9

1-iodo-3-methylpropane

(CH3)2CHCH2I

9.202

4.54 · 10-11

10 1-iodoethane

CH3CH2I

9.35

3.50 · 10-11

11 iodomethane

CH3I

9.538

1.74 · 10-11

12 d3- iodomethane

CD3I

9.538

2.01 · 10-11

13 1,1,1-trifluoro-2-iodoethane

CF3CH2I

9.998

1.04 · 10-11

14 1,1,1-trifluoro-2-fluoro-2iodoethane

CF3CHFI

10.32

6.05 · 10-12

15 trifluroiodomethane

CF3I

10.23

4.34 · 10-12

16 pentafluoroethyliodide

CF3CF2I

10.66

4.82 · 10-12

In cases where calibrated alkene spectra were not available identification of the alkene was based on the infrared spectra taken from the NIST infrared database. With the exceptions of iodomethane (CH3I) and neopentyl iodide ((CH3)3CCH2I) the yield of the alkene from the various alkyl iodides was substantial in every case and for some compounds approached, within the error limits, unit yield, e.g. iodoethane, 1,3-diodopropane and 1-iodo-2methylbutane. In every reaction investigated absorptions in the ir spectra were observed which could be attributed to formation of an alkane formed from the recombination of the alkyl radicals produced via abstraction of the I atom from the RI compound, i.e. RI + O o R + IO R + R + N2 o R-R + N2

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

203

0.50

absorbance (base 10)

0.40

0.30

d) c)

0.20

b) 0.10

a) 0.00 1300

1250

1200

1150

1100

1050

1000

950

900

850

800

750

700

-1

wavenumber (cm )

Figure 9.

Plot of the spectral data in the range 1300 – 700 cm-1 obtained in the irradiation of a 1-iodobutane (CH3CH2CH2CH2I)/NO2/N2 reaction mixture: a) spectrum of initial mixture before irradiation showing absorption bands from 1-iodobutane, b) the product spectrum after 10 minutes irradiation, c) difference spectrum of b) - a) with absorptions of 1-iodobutane zeroed, and d) reference spectrum of 1butene.

0.02

Intermediate formation of HOI

0.01

absorbance (base 10)

0.01 0.00 -0.01 -0.01 -0.02 -0.02

Decay of 1-iodobutane

-0.03 -0.03 1250

1150

1050

950

850

-1

wavenumber (cm )

Figure 10.

A difference spectra formed from the subtraction of a spectrum of a mixture of 1-iodobutane/NO2/N2 recorded before irradiation from one recorded after 4 minutes irradiation. The weak absorption with its centre at 1070 cm-1 is attributable to hypoiodous acid.

I. Barnes

204

The yields of the alkane products have not been quantified. The yield of the alkane will depend upon the experimental conditions used since the alkyl radicals can also react with the NO2 and NO in the system. Due to the presence of NOx in the system the formation of organic nitrates in small yields was also observed. It is interesting to compare the alkenes formed in the reactions of O(3P) with 1iodobutane and 2-iodobutane. In the reaction of O atoms with 1-iodobutane only one alkene can be formed, namely 1-butene, and this is observed with a yield of around 70%. In the case of 2-iodobtane, however, there is the possibility of the formation of both 1-butene and cis/trans -2-butene as the major alkene product (see Figure 11). The measurements show that the major alkene is 1-butene with a yield of around 46%. This results implies that adduct formation of the O atom between the I atom and the terminal methyl group is favoured over adduct formation between the O atom and the methylene group. Formation of trans-2-butene was observed, however, the yield was difficult to quantify and only a limit of <10% can be set at present. O H O + H3C

H

I

C

C

C

H

H

H

H H

H3C

H

I

C

C

C

H

H

H

CH3-CH2-CH=CH2

H

1-butene

+ HOI

1-iodobutane

hypoiodous acid

O H

I

H

C

C

C

H

H

H

I

H

H

C

C

C

H

H

H

major

H3C O + H3C

H

I

H

C

C

C

H

H

H

+ HOI

H O

2-iodobutane

H3C

Figure 11.

CH3-CH2-CH=CH2

H

H

minor

CH3-CH=CH-CH3 2-butene + HOI

Possible reaction pathways in the reaction of O(3P) with 1-iodobutane and 2iodobutane.

Conclusions The work has shown that a very small reactor can be used with a GC system to obtain kinetic data. Using the relative kinetic technique rate coefficients have been obtained for the reaction of O(3P) atoms with a series of alkyl iodides. Many of the rate coefficients are first time measurements. It has also been shown that alkenes and HOI are the major products of the reactions and the alkene has been quantified for the majority of the alkyl iodides studied. Acknowledgements This work was supported by the Bundesministerium für Bildung und Forschung within the “Atmosphärenforschung” programme (AFO 2000).

Kinetics, Products and Mechanism of O( 3P) Atom Reactions with Alkyl Iodides

205

References Barnes, I., Becker, K. H. and N. Mihalopoulos; An FTIR product study of the photooxidation of dimethyl disulfide, J. Atmos Chem., 18 (1994) 267-289. Barnes, I., K. H. Becker and T. Zhu; Near UV absorption spectra and photolysis products of difunctional organic nitrates, J. Atmos. Chem. 17 (1993) 353-373. Barnes, I., K. H. Becker, and S. Starke; FTIR spectroscopic observation of gaseous HOI, Chem. Phys. Lett. 196 (1992) 578-582. Carl, S. A. and J. N. Crowley; 298 K rate coefficients for the reaction of OH with i-C3H7I, n-C3H7I and C3H8, Atmos. Chem. Phys. 1 (2001) 1-7. Cavalli, F., H. Geiger, I. Barnes and K. H. Becker; FTIR kinetic, product, and modeling study of the OHinitiated oxidation of 1-butanol in air, Environ. Sci. Technol. 36 (2002) 1263-1270 Cotter, E. S. N., N. J. Booth, C. E. Canosa-Mas, D. J. Gray, D. E. Shallcross and R. P. Wayne; Reactions of Cl atoms with CH3I, C2H5I, 1-C3H7I, 2-C3H7I and CF3I: kinetics and atmospheric relevance, Phys. Chem. Chem. Phys., 3 (2001) 402–408. Cotter, S. A., C. E. Canosa-Mas, C. R. Manners, R. P. Wayne and D. E. Shallcross; Kinetic study of the reactions of OH with the simple alkyl iodides: CH3I, C2H5I, 1-C3H7I and 2-C3H7I, Atmos. Environ. 37 (2003) 11251133. Dillon, T. J. and D. E. Heard; A laser-induced fluorescence study of the O(3P) + CF3I and IO + NO reactions: Evidence for an association channel for O + CF3I at low temperatures, J. Photochem. Photobiol., 157 (2003) 223–230. Gilles, M. K., A.A. Turnipseed, R. K. Talukdar, Y. Rudich, P. W. Villalta, L. Gregory Huey, J. B. Burkholder and A. R. Ravishankara; Reactions of O(3P) with alkyl iodides: Rate coefficients and reaction products, J. Phys. Chem. 100 (1996) 14005-14015. Klaassen, J. J., J. Lindner and S. R.Leone; Observation of the Q1 OH(OD) stretch of HOI and DOI by Fourier transform infrared emission spectroscopy, J. Chem. Phys. 104 (1996) 7403-7411. Knyazev, V. D., V. S. Arutyunov and V. I. Vedeneev; The mechanism of O(3P) atom reactions with ethylene and other simple olefins, Int. J. Chem. Kinetics 24 (1992) 545-562. Loomis, R. A., J. J. Klaassen, J. Lindner, P. G. Christopher and S. R. Leone; Fourier transform infrared emission study of the mechanism and dynamics of HOI formed in the reaction of alkyl iodides with O(3P), J. Chem. Phys. 106 (1997) 3934-3947. Luo, D., J. A. Pierce, I. L. Malkina and W. P. L. Carter; Rate constants for the reactions of O(3P) with selected Monoterpenes, Int. J. Chem. Kinetics, 28 (1996) 1-8. Olariu, R. I., I. Barnes, K. H. Becker, and B. Klotz; Rate coefficients for the gas-phase reaction of OH radicals with selected dihydroxybenzenes and benzoquinones, Int. J. Chem. Kinet. 32 (2000) 696-702 Teruel, M. A., T. J. Dillon, A. Horowitz, and J. Crowley; Reaction of O(3P) with the alkyl iodides: CF3I, CH3I, CH2I2, C2H5I, 1-C3H7I and 2-C3H7I, Phys. Chem. Chem. Phys. 6 (2004) 2172-2178. Monks, P. S., L. J. Stief, D. C. Tardy, J. F. Liebman, Z. Zhang, S. –C. Kuo and R. B. Klemm; Discharge flowphotoionization mass spectrometric study of HOI: Photoionization efficiency spectrum and ionization energy, J. Phys Chem, 99 (1995) 16566-16570.

Kinetics of the Reaction between CF2 O and CH3OH Maximiliano A. Burgos Paci and Gustavo A. Argüello INFIQC, Departamento de Físico-Química, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, 5000 Córdoba, Argentina Key Words: Heterogeneous reactions, Fluorinated compounds, Kinetics, Methanol

Abstract The reaction between CH3OH and CF2O has been studied in the temperature range between 20 and 122 ºC. A considerable decrease in the reaction rate with increase in temperature was noted, indicating a contribution of heterogeneous character. Different types of surfaces and coatings have been used. The possible involvement of aerosols for the reaction to occur in the atmosphere is discussed. Introduction Methanol is a trace component of our atmosphere and significant concentrations (400-800 ppt) have been measured in the upper troposphere (Singh et al. 2000; Singh et al. 1995). A rough first inventory of the global sources of methanol indicates that its sources far exceed its known sinks (Singh et al. 2000). The large mismatch between the current atmospheric concentration and the estimated sources and sinks of this oxygenated species indicates that substantial and still unknown removal processes, other than removal with OH radicals, must exist. The existence of unknown, possibly heterogeneous processes, which could provide effective sinks for methanol, cannot be currently ruled out (Singh et al. 2000). Removal reactions for this species must, therefore, be found within the atmospheric constituents, and the search should not only be confined to homogeneous processes, but should also be extended to heterogeneous processes. We have been exploring many different atmospheric constituents as reaction partners for CH3OH in the gas phase using an in situ FTIR experimental setup to follow the course of the reactions. Among many very well-known substances, there is a species, whose concentration has been growing steadily in the atmosphere because it comes from the degradation of almost any fluorinated molecule released into the atmosphere, i.e. carbonyl fluoride (CF2O). CF2O could represent almost up to a third of the total inorganic fluorine in the stratosphere (Raper et al. 1987). We have found that CF2O reacts in the gas phase, though with a very small rate constant which makes the homogeneous process almost unimportant, in contrast the heterogeneous contribution of this reaction is very important. The relative importance of the heterogeneous contribution was studied using smallvolume reaction chambers, where wall effects were evident and allowed different surfaces and surface conditions to be investigated. We have previously demonstrated the active role played by the surface in the handling of fluorinated species (Pernice, 2002). CF2O, in particular, is considered one of the most stable end products in the atmospheric degradation of almost any CFC, HCFC, HFC or HFE because of its strong bonds that allow this molecule to survive most of the radiation found below 30 km of altitude without being degraded. Nevertheless, it reacts with

207 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 207–212. © 2006 Springer. Printed in the Netherlands.

208

M. A. Burgos Paci and G. A. Argüello

CH3OH, gives a unique product (the characterization of which is in progress) irrespective of the type of reaction (homogeneous or heterogeneous) and the rate of the reaction increases with decreasing temperature. This temperature behaviour is particularly interesting because the temperature profile of the atmosphere and the vertical distribution of CF2O favour (in the presence of particulate matter) the disappearance of methanol. Experimental Volatile compounds were manipulated in a conventional vacuum line with three U-traps, PTFS stop-cock valves (Young, London) and two pressure gauges (Bell and Howell). The line was connected directly to the reaction cell, which lies in the optical axis of an FTIR spectrometer (Bruker IFS28), in a configuration fairly similar to the one used in smog chamber experiments. This set-up allowed the timely recording of IR spectra and all the analyses were therefore performed by FTIR spectroscopy. The reactions have been carried out by mixing different pressures of CH3OH (1 – 30 mbar) and CF2O (1 – 25 mbar) in the reaction cell. A double-walled reaction cell connected to a thermostat was used to investigate the effect of temperature. Water and glycerine were used as heating liquids for temperatures between 20 – 90 and 90 – 122 ºC, respectively. In order to vary the type of surface the reactants were exposed to, three reaction cells were employed: a) a glass cell with and without small glass balls added, b) a stainless steel cell, and c) a glass cell with the inside walls covered with an inert wax (Halocarbon). In order to cover the glass cell with Halocarbon wax, it was first spread on the inside walls. The cell was then heated in an oven at 70 ºC for approximately 1 hour until the wax formed a film over the walls. Results and Discussion In order to follow the evolution of the system as the reaction takes place, IR spectra were recorded in all the experiments immediately after the reactants were mixed. Figure 1 shows IR spectra acquired before (A), after 15 min (B) and at the end (C) of the reaction of 18 mbar CH3OH and 3 mbar CF2O. The disappearance of CF2O was monitored using its PQR band at 1985 cm-1. The rate of disappearance was always faster than that of CH3OH. The time evolution of CH3OH was followed at 1095 cm-1 though it was quite difficult to quantify because of the interference of SiF4. The reaction results in the formation of a still unidentified product with characteristic IR bands at 1850 cm-1, which is typical of a C=O stretch, and other bands at 1250 cm-1 (C—F stretch), 1455 cm-1 (CH bending), 915 cm-1 and 785 cm-1 which are shown in panel (D) of Figure 1. Several experimental runs with different initial concentrations of each reactant were carried out to investigate the effect in the rate of product formation. The curves obtained were fitted in all cases with the equation [Prod] = Ao[1-exp(-k t)]. Plot A in Figure 2 shows the case where the initial pressure of CH3OH was fixed (8.0-9.0 mbar) and the CF2O initial pressure was set at 0.8; 10.0 and 11.0 mbar. It is worth noting that the amount of product formed increases as the initial pressure of CF2O is increased, i.e. it is much smaller for an initial pressure of 0.8 mbar than for one of 10.0 mbar. Another feature to notice is that when the initial pressure of CF2O is 10.0 – 11.0 mbar, and 0.8 mbar the stationary product concentration is reached at around 50 min. and 200 min. respectively. This indicates that the rate of the reaction increases as the initial

Kinetics of the Reaction between CF2 O and CH3 OH

209

pressure of CF2O increases. Analogous observations can be made when the initial pressure of CF2O is fixed and that of CH3OH is varied.

6

D

Abs(a.u.)

5 4

C

3

B

2 1

A

0 3000

2500

2000

cm Figure 1.

-1

1500

1000

500

Sequence of IR spectra obtained at different times of reaction.

The order of the rate with respect to CF2O was obtained by running the reaction with an excess of CH3OH and conversely, to obtain the order with respect to the concentration of CH3OH at least a ten fold excess of CF2O was used. Under these conditions the logarithm of the concentration of the limiting reagent as a function of time can be excellently fitted with a linear function, as shown in Plot B of Figure 2 for the case of CF2O. The order of the reaction is one with respect to each reagent and order two overall. Under the experimental conditions the pseudo first order constant for the disappearance of CF2O at 25 ºC was 2.5x10-3 s-1. Effect of temperature The temperature effect on the rate of reaction was studied between 20 and 122 ºC. Plot A of Figure 3 shows the formation of the main still unidentified product in three experiments performed at 25, 100 and 122 ºC with the same initial pressures of reactants. A decrease in the product formation rate was observed as the temperature increased. It is interesting to notice that at all temperatures the products are the same. The decrease of the rate at higher temperatures can be rationalized in terms of a heterogeneous reaction taking place on the surface of the cell, since as the temperature goes up there should be a correspondingly higher percentage of free gaseous molecules not adsorbed onto the walls that should react homogeneously.

M. A. Burgos Paci and G. A. Argüello

210

Presently, we are trying to determine whether the reaction could occur on different surfaces and whether it could give other products. 0.4

-0.5

A

0.3

B

-1.0

0.2

9mb CH3OH + 11.0mb CF2O 8mb CH3OH + 10.0mb CF2O 8mb CH3OH + 0.8mb CF2O

0.1

ln([CF2O])

Abs (a.u.)

-1.5 -2.0 -2.5 -3.0 -3.5 0.0

-4.0 0

50

100

150

time (min.)

Figure 2.

200

250

0

5

10

15

20

time (min)

A) Effect of the pressure of CF2O on product formation. B) Pseudo first order linear fit for the decomposition of CF2O.

Effect of different surfaces In order to investigate the effect of the surface with which the reactants collide, experiments were performed to change the surface of the cell. Plot B of Figure 3 shows the effect of using a glass (squares in the figure) or a stainless steel (circles) reaction vessel. As can be seen the type of surface clearly affects the reaction rate. Though the experimental runs did not reproduce exactly the same pressures, the differences in rate are beyond any reasonable doubt. It is observed that in the glass cell, the rate is faster than in the stainless steel cell. This observation can be taken as a proof of the catalytic effect of glass on the reaction. The ratio of the constants for the formation of product (taken from the equations [Prod] = Ao[1-exp(-k t)]) indicates that the rate is 20 times higher in the glass cell. A hypothesis could be that the OH radicals and silica groups on the glass surface facilitate the adsorption of CH3OH by dipole-dipole interactions thus providing the proper orientation of the molecule to be attacked or the weakening of the bond that will constitute the reaction coordinate. In the stainless steel cell there is no adsorption of CH3OH; thus the catalytic effect is reduced and the reaction should proceed almost entirely in the gas phase giving the same product as in the glass cell. Plot C in Figure 3 shows three experiments: one using a glass cell, another with a glass cell with glass beads added to increase the surface, and a third one employing a cell whose walls were covered with wax. From a comparison of the first two experiments (squares and triangles) it can be seen that the addition of glass surface does not have a quantitative effect on the reaction rate. This could be because the amount of glass in the walls of the cell far exceeds the active surface needed to catalyze the reaction and so the added glass balls do not increase the reaction rate. The diamonds in the figure correspond to the waxed glass cell. Here, an induction time is observed after which, the reaction proceeds as in the “normal” glass cell in terms of the amount of product formed. We believe that in the cell there may have been some points where the wax film was too thin and after the first 30 min of the reaction some molecules could have reached the

Kinetics of the Reaction between CF2 O and CH3 OH

211

glass surface; once this occurred the catalytic effect of glass accelerated the reaction. This reinforces the idea that a small amount of active surface is needed to catalyze the reaction. 0.40 0.35 0.30

Abs (a.u.)

0.25 0.20 0.15 0.10

A

0.05 0.00 0

10

20

30

40

50

60

time (min.)

0.40 0.5

0.35

0.4

0.25

Abs (a.u.)

Abs (a.u.)

0.30

0.20 0.15 0.10

0.2 0.1

0.05

B

0.00 0

60

120

180

240

time (min.)

Figure 3.

0.3

300

360

420

C

0.0 0

30

60

90

120

150

180

210

time (min.)

A) Temperature effect on the rate of product formation. Ŷ T=25 ºC; Ɣ T=100 ºC; Ÿ T=122 ºC. B) Product formation when the reaction is carried out in a glass (Ŷ) and a stainless steel (Ɣ) cell. C) Product formation when the reaction is carried out in a glass cell (Ŷ), a glass cell with glass beads (Ÿ) and a waxed cell (Ƈ).

Atmospheric Implications This preliminary study of the reaction between CF2O and CH3OH shows that it has an important heterogeneous component, giving a mean lifetime of 6 min. at 25 ºC for CH3OH when the reaction is carried out with an active surface. We think that the homogeneous gas phase reaction is too slow to be considered as an important process since it is 20 times slower than the heterogeneous reaction. Thus taking into account that: a) the reaction is catalyzed heterogeneously and in the atmosphere there is a high possibility of having heterogeneous processes due to the presence of aerosols, particulate matter, water droplets, etc., and b) the reaction rate increases as the temperature decreases and in the upper troposphere temperatures

212

M. A. Burgos Paci and G. A. Argüello

decrease to a mean value of -50 ºC. This reaction could have an important role in explaining the low concentration of CH3OH in the atmosphere. However, further work is needed to assess the feasibility of this hypothesis. Problems to solve are: i) the nature of the product of the reaction and its associated chemistry and photochemistry, ii) the stoichiometry of the reaction, and iii) the effect of atmospheric aerosols on the reaction rate. Acknowledgements The authors thank CONICET, SECyT and ANPCyT for financial support. M.A.B.P. thanks CONICET for a postdoctoral fellowship. Language assistance by Karina Plasencia is gratefully acknowledged. References Singh H., Y. Chen, A. Tabazadeh, Y. Fukui, I. Bey, R. Yantosca, D. Jacob, F. Arnold, K. Wohlfrom, E. Atlas, F. Flocke, D. R. Blake, N. Blake, B. Heikes, J. Snow, R. Talbot, G. Gregory, G. Sachse, S. Vay and Y. J. Kondo; Distribution and fate of selected oxygenated organic species in the troposphere and lower stratosphere over the Atlanatic, Geophys. Res. 105 (2000) 3795-3805. Singh H., M. Kanakidou P. J. Crutzen, and D. J. Jacob; High concentrations and photochemical fate of oxygenated hydrocarbons in the global troposphere, Nature 378 (1995) 50-54. Pernice H., H. Willner, K. L. Bierbrauer, M.Burgos Paci and G. A. Argüello; Fluoroformic Acid Anhydride, FC(O)OC(O)F, Angew. Chem. Int. Ed. 41 (20) (2002) 3832-3834. Raper O. F., C. B. Farmer, R. Zander and J. H. Park, Ground-Based Infrared Measurements of Tropospheric Source Gases Over Antarctica During the 1986 Austral Spring, J. Geophys. Res. 92 (1987) 9851-9856.

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System M. S. Chiappero1 , F. E. Malanca1 , G. A. Argüello1 S. Nishida , K. Takahashi2, Y. Matsumi2, M. D. Hurley3, and T. J. Wallington3 2

1

INFIQC, Departamento de Físico-Química, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, 5000 Córdoba, Argentina 2 Solar-Terrestrial Environment Laboratory and Graduate School of Science, Nagoya University, Toyokawa, Aichi 442-8507, Japan 3 Department of Physical and Environment Sciences, Ford Research and Advanced Engineering, SRL-3083, Ford Motor Company, Dearborn, MI 48121-2053, USA Key Words: Fluorinated peroxy radicals, Kinetic, Nitric Oxides, Organic nitrates, UV Spectroscopy

Abstract CF3Ox radicals are present in the atmosphere as oxidized species of a series of CFC´s, HFC´s and HCFC´s. They react with NOx species. In particular, the reaction between CF3O2 and NO which shows two reaction channels, CF3O2 + NO o CF3O + NO2

k1a

(1a)

CF3O2 + NO + M o CF3ONO2 + M

k1b

(1b)

has been revised using FTIR smog chamber techniques to provide the first systematic study of the nitrate yield thus allowing the determination of the branching ratio for both processes. In 700 Torr of N2/O2 diluent at 296 K the branching ratio is k1b/(k1a+k1b) = (1.67±0.27)×10-2 and the CF3ONO2 forming channel appears to approach the high-pressure limit for pressures above 100 Torr. This finding could probably provide more conventional energy transfer parameters to account for the pressure dependence in atmospheric models. CF3ONO2 is also the main product observed in the black lamp (Omax|355 nm) photolysis of pure CF3O2NO2. The thermal decomposition of the peroxide (which does not absorb significantly beyond 300 nm) leads to an equilibrium with CF3O2 and NO2 which does absorb and thus forms NO. Subsequent reactions lead to CF3O radicals, that in presence of NO2 form CF3ONO2, giving the overall reaction: CF3O2NO2 + hQ o CF3ONO2 + ½ O2

(2)

The low branching ratio in the formation of CF3ONO2, the photochemical characteristics of CF3O2NO2 just described and its probable even distribution in the free troposphere, led us to look closer at the UV features of the peroxide. The UV spectrum at room temperature was already known. We have determined the temperature dependence between 225 and 300 K. The results show an increase in the cross sections with temperature. From the data, photochemical half-lives in the atmosphere from 20 to 450 days result, depending on altitude and solar zenith angle. In terms of the process controlling CF3O2NO2 decomposition, calculations show that from the surface to around 2 km altitude, the thermal process is faster than photolysis. Beyond 2 km, photochemistry governs the decomposition for low and mid latitudes. At higher latitudes (50 degrees) the limiting process below the tropopause is thermal decay. 213 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 213–221. © 2006 Springer. Printed in the Netherlands.

M. S. Chiappero et al.

214 Introduction

In the atmosphere, the CFC´s, HFC´s and HCFC´s containing the CF3 moiety produce CF3Ox radicals that react with NOx. The reaction between CF3O2 and NO shows two possible channels: CF3O2 + NO o CF3O + NO2

(1a)

CF3O2 + NO + M o CF3ONO2 + M

(1b)

Channel (1a) converts NO into NO2 thereby leading to net O3 formation since the production of tropospheric ozone is due mainly to NO2 photolysis. NO2 + hQ (O < 420 nm) o NO + O(3P)

(3)

O(3P) + O2 + M o O3 + M

(4)

Reaction channel (1b) produces alkyl nitrates which are less reactive than either NO or NO2 and hence limits the photochemical formation of ozone. The formation of organic nitrate via reaction (1b) provides a tracer of photochemical ozone production in the atmosphere (Bertman et al., 1995; Flocke et al., 1991; Flocke et al., 1994; Roberts et al., 1998). Therefore, an understanding of the processes which control tropospheric ozone levels is essential in assessments of the global oxidizing capacity of the atmosphere and the anthropogenic contribution to global climate change. There are no literature data available concerning the branching ratio k1b/(k1a+k1b) and therefore no estimates of the importance of nitrate formation in reaction (1). Furthermore, there have been no direct studies of the nitrate yield for any halogenated peroxy radical. The organic nitrate is an inactive reservoir species for NOx. These reservoirs may be subject to dry and wet deposition (Treves et al., 2000), as well as thermal and/or photochemical decomposition. Peroxynitrates, RO2NO2, are thermally unstable and decompose to reform RO2 radicals and NO2. At typical temperatures of the lower troposphere the half life of peroxynitrates with respect to thermal decomposition is of the order of minutes whilst at the colder temperatures typical of the upper troposphere these compounds have half lives measured in months or years. Acting as reservoir compounds for NOx, they can transport NOx (and RO2 radicals) over long distances. The fluorinated organic nitrates CF3ONO2 and CF3O2NO2 have been well characterized (Sander et al., 2001; Kopitzky et al., 1998; Hohorst and DesMarteau, 1974) and show UV absorption below 300 nm. CF3O2NO2 absorbs significantly in the wavelength region 200-300 nm (Kopitzky et al., 1998). Photolysis is thus a potential loss mechanism for CF3O2NO2 in the stratosphere which needs to be considered. UV absorption cross section measurements as a function of temperature are also needed as inputs to global atmospheric chemistry programs in order to calculate photolysis frequencies. The majority of such currently available measurements are related to primary pollutants such as the CFCs (Hubrich et al., 1977; Simon et al., 1988; Merienne et al., 1990); HCFCs (Robbins, 1976; Vanlaethem-Meuree et al., 1978); HFEs (Orkin et al., 1999); and alkyl iodides (Roehl et al., 1997; Rattigan et al., 1997; Fahr et al., 1995). There have been comparatively few studies of the UV spectra of stable primary oxidation products of halogenated compounds (e.g., CF3C(O)Cl, CF3C(O)F, and CF3C(O)H, (Rattigan et al., 1993)).

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System

215

The data base for thermally unstable products such as CF3C(O)O2NO2 (perfluoroperoxyacetylnitrate) (Libuda et al., 1995), and CF3O2NO2 is particularly sparse. For the reasons referred to above, the aims of this work are (i) the study of the branching ratio k1b/(k1a+k1b) to elucidate the factors influencing the CF3ONO2 yield in the reaction of CF3O2 radicals with NO, (ii) the study of the photochemical (O>300 nm) behaviour of CF3O2NO2 and (iii) the study of the influence of temperature on the UV absorption features of CF3O2NO2. Experimental

The study of the reaction between CF3O2 and NO at 296 ± 2 K, was carried out with different mixtures of CF3N2CF3 (6.18–11.8 mTorr), NO (14.7–241 mTorr) and O2 (19.2–100 Torr ) in N2 diluent (20–700 Torr) that were admitted into a 140 L Pyrex reactor interfaced to a Mattson Sirus 100 FTIR spectrometer. The reactor was surrounded by 22 fluorescent blacklamps (Wallington and Japar, 1989). Each mixture was subjected to UV irradiation and the time evolution of the system was followed by FTIR spectroscopy (Wallington and Ball, 1995a). CF3O2 radicals were produced in situ by the reaction of O2 with the CF3 radicals formed by photolyis of CF3N2CF3 (Omax<355 nm). Control experiments were performed in which CF3N2CF3/NO/O2/N2 mixtures as well as their products were allowed to stand in the dark in the chamber for 30 min. In other experiments, the effect of black lamp radiation (GE FL6 BLB) on CF3O2NO2 was studied using a 23 cm long quartz cell fitted at both ends with KBr windows. The quartz cell was mounted in the sample compartment of a Bruker IFS28 FTIR spectrometer (Bruker, Karlsruhe, Germany) equipped with a DTGS detector and KBr beam splitter. All experiments were performed at 298 K. Spectra were obtained by co-adding 8 interferograms in the range 4000 to 400 cm-1 with 2 cm-1 resolution. In addition, the dependence of the UV absorption with temperature was studied using a diode array UV spectrometer with a standard 10 cm quartz gas cell mounted and aligned inside an evacuated metal box with two opposed quartz windows. The body of the cell was surrounded with a plastic coil connected to an external thermostat-cryostat for temperature control (300-220 K). The temperature of the gas within the cell could be regulated to within 0.1oC. For atmospheric half-life calculations, the program GCSOLAR release 1.2 (Zeep et al.,1997) with GCSOLAR ELEVATION (which allows to compute photolysis rate constants of a particular substance as function of elevation above sea level), was used. Unity quantum yield, a terrestrial atmosphere and Greenwich meridian, were assumed for all latitudes. Reagents The reagents CF3N2CF3, CF3O2NO2, and CF3ONO2 were obtained as described below: CF3N2CF3 was synthesized by passing cyanogen chloride diluted with nitrogen or helium, over a bed of silver fluoride (AgF2) and collecting the product in a cold trap (pentane/liquid nitrogen or liquid nitrogen) as described elsewhere (Wallington and Ball, 1995b).

216

M. S. Chiappero et al.

CF3O2NO2 was synthesized by mixing (CF3CO)2O, NO, and O2 in a 12 L reactor, maintained at -25 ºC and subjecting the mixture to 254 nm radiation (Burgos Paci et al., 2003). CF3O2NO2 was purified by fractional distillation under vacuum at -100/-120/-196 ºC. The peroxide was obtained in the -120 ºC fraction (Chiappero et al., 2004). CF3ONO2 was synthesized photolysing a pure sample of CF3O2NO2 with a black lamp (Omax|355 nm). The product was purified by trap to trap distillation (Chiappero et al, 2004). Other reagents (N2, O2, and NO) were obtained from commercial sources at purities >99% and were used as received. Results and Discussion The photolysis of CF3N2CF3 in the presence of O2 leads, through the sequence: CF3N2CF3 + hQ o 2 CF3 + N2

(5)

CF3 + O2 + M o CF3O2 + M

(6)

to the formation of CF3O2 radicals. In the presence of excess NO (14.7-241 mTorr) the fate of CF3O2 radicals is dominated by reaction (1). COF2 and CF3ONO2 were the only observed carbon-containing products. Within the experimental uncertainties the observed formation of COF2 and CF3ONO2 accounts for 100% of the loss of the CF3N2CF3. The observed formation of CF3ONO2 provides information concerning k1b/(k1a+k1b). Assuming that reaction (1) is the sole loss of CF3O radicals then the branching ratio is k1a/k1b ='[COF2]/'[CF3ONO2].

' [COF2] (mTorr)

10

5 Total Total Total Total Total

0 0.0

pressure pressure pressure pressure pressure

0.1

= = = = =

20 Torr 50 Torr 100 Torr 300 Torr 700 Torr

0.2

' [CF 3 ONO 2 ] (mTorr)

Figure 1.

Formation of COF2 versus CF3ONO2 for experiments conducted at selected pressures. The lines plotted correspond to linear least square fits. The solid line comprises the data points for 100, 300, and 700 Torr and has a slope = 0.0167 (see text for details).

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System

217

As the branching ratio k1b/(k1a+k1b) is expected to display a dependence on total pressure due to the termolecular nature of reaction (1b), experiments were carried out at different total pressures (N2/O2 as diluent). As seen from Figure 1, within the admittedly significant data scatter there is no discernable effect of total pressure over the range 100–700 Torr on the relative yields of COF2 and CF3ONO2. However, the [COF2]/[CF3ONO2] ratio observed at 20 and 50 Torr total pressure is significantly different from that at 100–700 Torr. The ratio k1b/(k1a+k1b) has values of 3.5 ×10-3, 7.6 ×10-3 and 1.67 ×10-2 for total N2/O2 diluent pressures of 20, 50 and 100-700 Torr and 296 K, respectively. The simplest explanation of these observations is that the rate of reaction (1b) is close to its high-pressure limiting value for pressures above 100 Torr but in the fall off regime at pressures of 50 Torr and below. It is likely that the pressure dependence reported herein for reaction (1b) can be modelled using more conventional energy transfer parameters as Barker and Golden (2003) suggest. The effect of the UV radiation (Omax|355 nm) over a room temperature equilibrated CF3O2NO2/NO2 mixture (CF3O2NO2 2.6 mbar) can be seen in Figure 2. This figure shows IR spectra before and after UV irradiation. IR features attributable to CF3O2NO2 at 792, 1192, 1244, 1303, and 1762 cm-1 (Kopitzky et al., 1998) are visible in panel A.

2 .0 1 .5 1 .0

A : b e fo re U V

0 .5 0 .0

Absorbance

1 .5 1 .0

B : 1 5 m in u te s U V

0 .5 0 .0

1 .5 1 .0

C : 3 0 m in u te s U V

0 .5 0 .0

1 .5 1 .0

D : 4 0 m in u te s U V

0 .5 0 .0

600

800

1000

1200

1400

1600

1800

2000

W a v e n u m b e r (c m -1 )

Figure 2.

IR spectra acquired before (A) and after (B), (C) and (D) UV irradiation. It can be seen the decrease in CF3O2NO2 concentration and the appearance of peaks at 788, 1156, 1243, 1265 and 1748 cm-1 which can be assigned to CF3 ONO2 (Sander et al., 2001). The bands at 1944 cm-1 correspond to COF2.

The results could be interpreted as follows: in the mixture CF3O2NO2/NO2, the peroxide (which does not absorb significantly beyond 300 nm) (Kopitzky et al., 1998) exists in equilibrium with CF3O2 radicals and NO2 CF3O2NO2 + M o CF3O2 + NO2 + M

(7)

CF3O2 + NO2 + M o CF3O2NO2 + M

(-7)

218

M. S. Chiappero et al.

and the irradiation of the mixture converts NO2 into NO (reaction 3) (Crutzen, 1973), which is then oxidized back to NO2 via reaction with CF3O2 radicals (reaction 1a). Subsequently, the resulting CF3O radicals react with NO2 to give CF3ONO2 CF3O + NO2 + M o CF3ONO2 + M

(8)

The result is photochemical conversion of CF3O2NO2 into CF3ONO2 in high yield with an overall stoichiometry of CF3O2NO2 o CF3ONO2 + ½ O2

(2)

The perfluoromethyl peroxinitrate has a dissociative UV spectrum below 300 nm similar to those of other peroxynitrates. However, as noted previously (Kopitzky et al., 1998), it is approximately a factor of 2 less intense (Morel et al., 1980). Table 1 shows the temperature dependence of the UV cross sections of CF3O2NO2 at 233, 253, 273, 294 and 300 K. As the peroxynitrate exists in equilibrium with NO2, and this species in turn exists in equilibrium with N2O4, at lower temperatures a greater fraction of the NO2 exists in the dimer form N2O4. NO2 and N2O4 absorb in the wavelength range studied (N2O4 being a particularly strong absorber) and the features of these two substances must be subtracted from the experimental spectra. This stripping process was repeated with the measurements at low temperature taking into account the variation in concentrations due to the change in the value of the equilibrium constant. The increase in absorption cross section with temperature seen from examination of Table 1 is consistent with an increased population of the vibrational and rotational levels of the ground electronic state of the molecule at higher temperatures (Okabe, 1978). In the troposphere, the temperature decreases with altitude increase up to the tropopause (approximately 10 km altitude). In this zone the temperature is close to 220 K and precisely, under these conditions, the peroxynitrate UV absorption cross sections are the least. This means that its photodissociation rate will be lower than at the surface. The precise value integrated for the whole absorption region will depend, of course, on the solar flux, that is richer in short wavelength radiation than the surface flux. In fact, a detailed calculation of the CF3O2NO2 half life and its dependence with altitude and latitude shows a maximum half life at around 15 km. This maximum reaches values as long as 220 days for mid-latitudes allowing the compound to be transported though the atmosphere. Therefore, CF3O2NO2 could be transported either through convective stream to the stratosphere, or to remote tropospheric places where it can dissociate again. From the surface up to 2 km, the thermal decomposition loss (Mayer-Figge et al., 1996) is always faster than the photochemical loss. Beyond this height the trend is opposite at low and mid latitudes and photochemistry governs the loss. At high latitudes, the photochemical half life is again longer. The use of Kopitzky´s cross sections, with no temperature dependence, results in a steady decrease of the half life with altitude. Furthermore, the absolute values are all shorter than 20 days; therefore our temperature dependent cross sections could help to better understand the fate of CF3O2NO2 since the half life is in any case considerablely longer than thought before.

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System Table 1.

219

CF3O2NO2 cross-sections (in units of 10-20 cm2 molecule-1) at different temperatures. Values are averaged over 5 nm intervals. Values printed in italics are derived from absorbance readings close to the baseline drift, therefore, their uncertainties are high. CF3O2NO2 cross-sections

 O (nm) 200 205 210 215 220 225 230 235 240 245 250 255 260 265 270 275 280 285 290 295 300 305 310 315 320 325 330 335 340

Kopitzky 301 K 220 150 120 74 54 46 38 32 27 22 18 14 11 8.2 5.8 4.0 2.6 1.7 1.0 0.69 0.45 0.28 0.20 0.13 0.08 0.05 0.037 0.024 0.014

300 K 215 149 115 77 58 47 40 35 29.5 24.0 19.0 14.7 11.5 8.5 5.8 3.96 2.76 1.85 1.31 0.88 0.60 0.45 0.31 0.23 0.16 0.12 0.096 0.040 0.029

294 K 215 149 111 76 58 47 40 35 29.3 23.8 18.9 14.8 11.2 8.2 5.8 3.94 2.65 1.75 1.18 0.76 0.5 0.35 0.23 0.16 0.11 0.071 0.048 0.033 0.022

This Work 273 K 216 150 114 77 57 47 39 33 27.7 22.3 17.6 13.6 10.1 7.3 5.1 3.32 2.15 1.35 0.86 0.54 0.32 0.22 0.14 0.086 0.054 0.034 0.022 0.014 0.009

253 K 216 150 115 78 56 46 38 32 26.6 21.3 16.8 12.8 9.5 6.8 4.6 2.96 1.84 1.11 0.66 0.4 0.22 0.14 0.085 0.051 0.03 0.018 0.011 0.007 0.004

233 K 215 149 114 77 55 45 37 30 25.7 20.5 16.2 12.3 9.1 6.4 4.3 2.68 1.57 0.87 0.45 0.23 0.11 0.058 0.03 0.015 0.008 0.003 0.001 -

Conclusions FTIR smog chamber techniques have been used successfully to show that reaction of CF3O2 radicals with NO proceeds via two channels giving either CF3O + NO2 (1a) or CF3ONO2 (1b) as products. The branching ratio for CF3ONO2 formation has been determined to be, k1b/(k1a+k1b) = (1.67 ± 0.27) × 10

-2

in 700 Torr of N2/O2 diluent at 296 K. Reaction (1b) appears to be close to its high pressure limiting value for pressures above 100 Torr.

220

M. S. Chiappero et al.

CF3O2NO2 reacts both by photochemical and thermal processes. The photolysis of a mixture in equilibrium of CF3O2NO2 at Omax|355 nm radiation leads to the formation of CF3ONO2 due mainly to the photolysis of NO2. There is a clear dependence of the cross sections with temperature. Although the photochemical half-life increases with latitude, the decomposition is mainly thermal within the boundary layer. If the temperature dependence in the cross sections were not taken into account, calculations would predict a half-life governed solely by photolysis. Acknowledgements T.J.W. and M.S.C. thank the National Science Foundation for a WISC grant. The Nagoya group thanks both Grants-in-Aid from the Ministry of Education, Science, and Culture Financial and also the Sumito Foundation for financial support. M.S.C., F.E.M. and GAA thank to CONICET, SeCyT-UNC, Antorchas for financial support. References Barker J.R. and D.M. Golden; Master Equation Analysis of Pressure-Dependent Atmospheric Reactions, Chem. Rev. 103 (2003) 4577-4592. Bertman S.B., J.M. Roberts and H. Westberg; Evolution of alkyl nitrates with air mass age, J. Geophys. Res., 100 (1995) 22805-22814. Chiappero M.S., M.A. Burgos Paci, G.A. Argüello and Wallington T.J.; Novel synthesis of trifluoromethylnitrate, CF3ONO2, Inorg. Chem. 43, (2004) 2714-2716. Crutzen P.J.; A discussion of the chemistry of some minor constituents in the stratosphere and troposphere, Pure Appl. Geophys. 106 (1973) 1385-1399. Fahr A., A.K. Nayak and R.E. Huie; Temperature dependence of the ultraviolet absorption cross section of CF3I, Chem. Phys. 199 (1995) 275-284. Flocke F., A. Volz-Thomas and D. Kley; Measurements of alkyl nitrates in rural and polluted air masses, Atmos. Environ. 25A (1991) 1951-1960. Flocke, F., A. Volz-Thomas and D. Kley; The use of alkyl nitrate measurements for the characterization of the ozone balance at TOR-Station N. 11, Schauinsland, in: P. B. Borrell, P. M. Borrell, T. Cvitas, and W. Seiler (eds), Transport and transformation of pollutants in the troposphere - EUROTRAC Symposium, SPB Academic Publishing (1994) 243-247. Hohorst F.A. and D.D. DesMarteau; Some reactions of CF3OO derivatives with inorganic nitrogen compounds. Synthesis and vibrational spectrum of trifluoromethyl peroxynitrate, Inorg. Chem. 13 (1974) 715-719. Hubrich C., C. Zetsch and F. Stuhl; Absorptionsspektren von halogenierten methanen im bereich von 275 bis 160 nm bei temperaturen von 298 und 208 K, Ber. Bunsenges. 81 (1977) 437- 442. Kopitzky R., H. Willner, H.-G. Mack, A. Pfeiffer and H. Oberhammer; IR and UV absorption cross sections, vibrational analysis, and the molecular structure of trifluoromethyl peroxynitrate, CF3OONO2, Inorg. Chem. 37 (1998) 6208-6213. Libuda H.G. and F. Zabel; UV absorption cross sections of acetyl peroxynitrate and trifluoroacetyl peroxynitrate at 298 K, Ber. Bunsenges. Phys. Chem. 99 (1995) 1205-1213. Merienne M.F., B. Coquart and A. Jenouvrier; Temperature effect on the ultraviolet absorption of CFCl3, CF2Cl2 and N2O, Planet. Space Sci. 38 (1990) 617-625. Mayer-Figge A., F. Zabel and K.H. Becker; Thermal decomposition of CF3O2NO2, J. Phys. Chem. 100 (1996) 6587-6593. Morel O., R. Simonaitis and J. Heicklen; Ultraviolet absorption spectra of HO2NO2, CCl3O2NO2, CCl2FO2NO2, and CH3O2NO2, Chem. Phys. Lett. 73 (1980) 38-42.

New Kinetic and Spectroscopic Measurements in the CF3Ox + NOx System

221

Okabe H.; Photochemistry of small molecules; (1978) p.p. 18-20. John Wiley & Sons, New York. Orkin V.L., E. Villenave, R.E. Huie and M.J. Kurylo; Atmospheric lifetimes and global warming potentials of hydrofluoroethers: Reactivity toward OH, UV spectra, and IR absorption cross sections, J. Phys. Chem., A. 103 (1999) 9770-9779. Rattigan O.V., O. Wild, R.L. Jones and R.A. Cox; Temperature-dependent absorption cross-sections of CF3COCl, CF3COF, CH3COF, CCl3CHO and CF3COOH, J. Photochem. Photobiol. A: Chem. 73 (1993) 1-9. Rattigan O.V., D.E. Shallcross and R.A. Cox; UV absorption cross-sections and atmospheric photolysis rate of CF3I, CH3I, C2H5I and CH2Icl, J. Chem. Soc. Faraday Trans. 93 (1997) 2839-2846. Roberts J. M., S.B. Bertman, D.D. Parish, F.C. Fehsenfeld, B.T. Jobson and H. Niki; Measurement of alkyl nitrates at Chebogue Point, Nova Scotia during the 1993 North Atlantic Regional Experiment (NARE) intensive, J. Geophys. Res., 103 (1998) 13569-13580. Robbins D.E.; Photodissociation of methyl chloride and methyl bromide in the atmosphere, Geophys. Res. Lett. 3 (1976) 213-216. Sander S., H. Willner, H. Oberhammer and G.A. Argüello; Trifluormethylnitrat, CF3ONO2, Z. Anorg. Allg. Chem. 627 (2001) 655-661. Simon P.C., D. Gillotay, N. Vanlaethem-Meuree and J. Wisemberg; Ultraviolet absorption cross-sections of chloro and chlorofluoro-methanes at stratospheric temperatures, Atmos. Chem. 7 (1988) 107-135. Treves K., L. Shragina and Y. Rudich; Henry’s law constants of some E-, J-, and G-hydroxyalkyl nitrates of atmospheric interest, Environ. Sci. Technol. 34 (2000)1197-1203. Vanlaethem-Meuree N., J. Wisemberg and P.C. Simon; Influence de la temperature sur les sections efficases d'absorption des chlorofluoromethanes dans l'ultraviolet, Bull. Acad. Roy. Belgique, Cl. Sci. 64 (1978) 31. Wallington T.J. and S.M. Japar; Fourier Transform Infrared Kinetic Studies of the Reaction of HONO with HNO3, NO3, and N2O5 at 295K, J. Atmos. Chem., 9 (1989) 399-409. Wallington, T.J. and Ball, J.C.; Atmospheric chemistry of CF3O radicals: reaction with O3. Chem. Phys. Lett. 234 (1995a) 187-194. Wallington T. J. and J.C. Ball; Atmospheric chemistry of CF3O radicals: reaction with CH4, CD4, CH3F, CF3H, 13 CO, C2H5F, C2D6, C2H6, CH3OH, i-C4H8, and C2H2, J. Phys. Chem. 99 (1995) 3201-3205. Zeep R.G. and D.M. Cline; Rates of Direct Photolysis in Aquatic Environment, Environ. Sci. and Technol. 11 (1997) 359-366.

Kinetic Study of the Temperature Dependence of the OH Initiated Oxidation of Dimethyl Sulphide Mihaela Albu1,2, Ian Barnes1, and Raluca Mocanu 2 1

Bergische Universität Wuppertal, Physikalische Chemie/FB C, Gaußstraße 20, D-42097 Wuppertal, Germany 2 ”Al.I. Cuza” University of Iasi, Faculty of Chemistry, Department of Analytical Chemistry, Carol I Boulevard 11, 700506 Iasi, Romania Key Words: Dimethyl sulphide, Hydroxy radical, Kinetics, Marine atmosphere, Photooxidation

Introduction Dimethyl sulphide (CH3SCH3, DMS) is the dominant natural sulphur compound emitted from the world’s oceans (Berresheim et al., 1995, Urbanski and Wine, 1999), accounting for about one quarter of global sulphur gas emissions. Oceanic DMS, through its oxidation products, is proposed to play a key role in climate regulation, especially in the remote marine atmosphere (Charlson et al., 1987). At the low NOx levels that characterize the remote marine boundary layer, the reaction with OH appears to be the most important loss process for DMS: OH + DMS ĺ products

(1)

The kinetics of the reaction of OH radicals with DMS have been extensively studied by a variety of direct and competitive techniques. The three early studies: two flashphotolysis/resonance fluorescence studies (in Ar/SF6 buffer gases) by Atkinson et al. (1978) and Kurylo (1978) and a competitive rate study (in 760 Torr of air) by Cox and Sheppard (1980) agreed that the rate coefficient is approximately 9 × 10-12 cm3 molecule-1 s-1 at ambient temperature. Atkinson et al. and Kurylo reported a small negative activation energy for the reaction. In addition, Atkinson et al. suggested that the mechanism of the reaction was a hydrogen abstraction. These three studies seemed to indicate that the rate coefficient was well defined, and unaffected by the presence of O2. However, in a later flash-photolysis/resonance fluorescence study Wine et al. (1981) obtained a rate coefficient of 4.3 × 10-12 cm3 molecule-1 s-1 and observed a small positive activation energy for the reaction. The authors suggested that in the previous studies, impurities in the DMS samples used may have affected the results. Also, the results of Cox and Sheppard could have been affected by secondary reactions leading to the loss of DMS. MacLeod et al. (1984) measured the rate coefficient at 373 and 573 K (no air present) and a value of 10.4 × 10-12 cm3 molecule-1 s-1 at room temperature was obtained from their data. They also reported a small positive activation energy. Atkinson et al. (1984) reinvestigated the reaction using a relative rate technique, and reported a rate coefficient of 10.3 × 10-12 cm3 molecule-1 s-1 at room temperature in the presence of 760 Torr of air. This was in reasonable agreement with Cox and Sheppard’s value in air, and the first two studies in the absence of air, but was not in agreement with the value of Wine et al. At this time, the situation was confused. But, numerous later studies: Martin et al. (1985), Wallington et al. (1986), Hynes et al. (1986), Hsu et al. (1987), Barnes et al. (1988), Nielsen et al. (1989), Abbatt et al. (1992), Hynes et al. (1995), Barone et al. (1996), and Williams et al. (2001) were carried out. The value of the activation energy was not satisfactorily resolved by these studies: Martin et al. and Wallington et al. reported small negative values for activation energy, whereas Hynes et al. (1986) and Hsu et al. reported small positive values. 223 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 223–230. © 2006 Springer. Printed in the Netherlands.

224

M. Albu et al.

Several authors have shown that rate coefficient was affected by the presence of O2 (Wallington et al. (1986), Hynes et al. (1986), Barnes et al. (1988)). In their comprehensive work, Hynes et al. (1986) found that the effective rate coefficient for OH + DMS and its deuterated analog (DMS-d6) was dependent on oxygen concentration, increasing as the partial pressure of oxygen was increased. The “oxygen enhancement” was the same for both DMS and its deuterated analog, showing no kinetic isotope but a significant negative temperature dependence. The experimental kinetic observations for reaction of OH radicals with DMS and DMS-d6 are consistent with a two-channel reaction mechanism involving both hydrogen abstraction (O2-independent channel) and reversible adduct formation in competition with adduct reaction with O2 (O2-dependent channel): OH + CH3SCH3 ĺ CH3SCH2 + H2O OH + CH3SCH3 (+M) ļ CH3S(OH)CH3 + (M)

(1a) (1b, 1-b)

CH3S(OH)CH3 + O2 ĺ products

(1c)

OH + CD3SCD3 ĺ CD3SCD2 + HDO

(2a)

OH + CD3SCD3 (+M) ļ CD3S(OH)CD3 + (M)

(2b, 2-b)

CD3S(OH)CD3 + O2 ĺ products

(2c)

In the work of Hynes et al. (1986) the temperature dependence of the rate coefficient for DMS + OH in 760 Torr air, together with temperature dependence of the branching ratio were determined. The authors obtained the following expression for kobs for one atmosphere of air: kobs = {T exp(-234/T) + 8.46 x 10-10 exp(7230/T) + 2.68 x 10-10 exp(7810/T) }/ {1.04 x 1011 T + 88.1 exp(7460/T)} The branching ratio for abstraction being given by: k1a/kobs = 9.6 x 10-12 exp(-234/T) /kobs. For many years, the two-channel mechanism has been widely accepted as the operative mechanism for the OH-initiated oxidation of DMS, and the expressions for the rate coefficient and inferred branching ratio recommended as the best currently available for mechanistic interpretation and atmospheric modelling purposes. A recent re-evaluation of the rate coefficient and the branching ratio has been made by Williams et al. (2001) using the pulsed laser photolysis-pulsed laser induced fluorescence (PLP-PLIF) technique. The effective rate coefficient for the reaction of OH + DMS and OH + DMS-d6 was determined as a function of O2 partial pressure at 600 Torr total pressure in N2/O2 mixtures; the temperature was 240 K for DMS and 240, 261, and 298 K for DMS-d6. This new work shows that at low temperatures the currently recommended expression underestimates both the effective rate coefficient for the reaction and also the branching ratio between addition and abstraction. For example, at 261 K a branching ratio of 3.6 was obtained as opposed to a value of 2.8 based on the work of Hynes et al. (1986). At 240 K the discrepancy increases between a measured value of 7.8 and a value of 3.9 using the extrapolated values from the 1986 work of Hynes et al. (the branching ratio is defined here as (kobs-k1a)/k1a). In addition, at 240 K the expression for kobs in 1 atm air based on the work of Hynes et al. (1986) predicts a value which is a factor of 2 lower then the value measured at

Kinetic Study of the Temperature Dependence of the OH Initiated Oxidation

225

600 Torr. Because of the size of these discrepancies at low temperatures, further kinetic studies are necessary to validate the results of Williams et al. (2001). The preferred value for the rate coefficient for reaction (1a) from the review of Atkinson et al. (2004) is k1a = 1.13 x 10-11 exp(-253/T) cm3 molecule-1 s-1 over the temperature range 240 - 400 K (k1a = 4.8 x 10-12 cm3 molecule-1 s-1 at 298 K). This value is based on the studies of Wine et al. (1981), Hynes et al. (1986), Hsu et al. (1987), Abbatt et al. (1992), and Barone et al. (1996). The preferred value for the rate coefficient for reaction (1b) from the review of Atkinson et al. (2004) is k1b = 1.0 x 10-39 [O2] exp(5820/T)/{1 + 5.0 x 10-30 [O2] exp(6280/T)} cm3 molecule-1 s-1 over the temperature range 240 - 360 K (k1b= 1.7 x 1012 cm3 molecule-1 s-1 at 298 K, 1 bar air). This value is based on the studies of Hynes et al. (1986), and Williams et al. (2001). Experimental The rate coefficients for the gas-phase reaction of OH radicals with dimethyl sulphide were determined using a relative rate method. The kinetic experiments were performed at a total pressure of 1000 mbar diluent gas (N2, synthetic air, or N2/O2 mixtures), at three different O2 partial pressure (~0 mbar, 205 mbar, and 500 mbar) and six different temperatures (250 K, 260 K, 270 K, 280 K, 290 K, and 299 K). The photolysis of hydrogen peroxide (H2O2) was used as the OH radical source: H2O2 + hȞ (Ȝ < 370 nm) ĺ 2 OH Ethene, propene, and iso-butene have been used as the reference organic compounds. The OH radical rate coefficients of these compounds are reliably established (Atkinson et al., 1997). All the experiments were carried out in a 336 l reaction chamber, which is outlined schematicly in Figure 1.

Figure 1.

Schematic diagram of the 336 l reaction chamber.

226

M. Albu et al.

The reactor is composed of 2 borosilicate glass cylinders of length 150 cm and different diameters which are mounted concentrically one within the other. The inner tube has a diameter of 60 cm. Both ends of the cylinders are closed with stainless steel flanges. Elastic KEL-F-seals (GDF Bauart 314) have been used as the sealing material in the entire reactor. The volume of the reactor after deduction of the volume for various fittings is 336 l. The reactor can be temperature controlled to within ±1 in the range 230 to 300 K. Isododecane (Maloterm P1), which is used as the coolant, is circulated continuously through the cavity between the two glass cylinders. The isododecane is cooled by three cryostats (HAAKE N3) connected in series. The temperature of the coolant in the cavity is controlled continuously by 8 platinum resistance thermometers. The reactor is isolated by several layers of the elastomeric nitrile rubber material Armaflex and the entire system is encased in a rectangular aluminium housing. The space between the outer housing and the end flanges is flushed with dry air to avoid condensation of water. Two other quartz glass tubes with diameters of 230 and 244 mm are mounted (concentrically one within the other) and centrally within the inner main reactor cylinder. The smaller of these central tubes contains 12 photolysis lamps. The lamps are of 2 types: 6 Philips TLA 40W (300 ” Ȝ ” 450 nm Ȝmax = 360 nm) and 6 Philips TUV 40W (main emission at Ȝ = 254 nm). The area between the glass cylinders can be cooled either by flowing coolant or dry air. The reactor is equipped with 2 White mirrors systems, one for UV spectroscopy and one for FT-IR spectroscopy. The mirror systems are mounted on the end flanges. The in and out coupling of the probing light is through windows (KCl for the IR and quartz for the UV) mounted on the flange containing the field mirror of the White system. The adjacent flange has connectors for the pumping system, pressure gauges and gas bottles. In addition an easily accessible smaller flange is mounted on the main flange for checking the condition of the reactor interior or placing samples inside the reactor. This small flange also contains a window for visual inspection of samples or chemical gas mixtures during measurements. The spectrometers and the outer reactor housing are connected with one another by a plastic housing, which is flushed with dry air and through which the analysing light beams pass. The starting concentrations (calculated at the working temperature) of DMS and of the reference organic compounds were in the range 1 to 3 ppm and that of H2O2 was approximately 20 ppm. After introducing the reactants into the chamber, the reaction chamber was filled to 1000 mbar with nitrogen, synthetic air, or nitrogen/oxygen mixtures in accordance with the type of experiment which was performed. The concentration-time behaviours of DMS and the reference organic compounds were followed over 30 - 35 min time periods by FT-IR spectroscopy. After acquiring 10 - 15 spectra of the mixture in dark (to control that no dark reaction takes place and to follow possible wall losses), the photooxidation was initiated by switching on the lamps (6 Philips TLA 40W) and then 15 - 20 spectra were recorded. Spectra were obtained by co-adding 64 scans which yielded a time resolution of 1 min. All the reactants were used without further purification. Dimethyl sulphide was supplied by Sigma Aldrich with a stated purity of 99%. H2O2 (85% wt) was supplied by Peroxide Chemie GmbH. The reference hydrocarbons: ethene, propene, and iso-buthene, were supplied by Messer-Griesheim with stated purities of 99.95%, 99.95%, and • 99%, respectively. The following gases were supplied by Messer-Griesheim: nitrogen (99.999%),

Kinetic Study of the Temperature Dependence of the OH Initiated Oxidation

227

oxygen (99.999%), and hydrocarbon free synthetic air (99.999%) with a N2:O2 ratio of 20.5%:79.5%. Results and Discussions The kinetic data obtained in this work shows that, for all the temperatures studied, the rate coefficient for DMS + OH is dependent on the O2 partial pressure, increasing as the O2 partial pressure is increased from ~0 to 500 mbar. At a fixed O2 partial pressure, the rate coefficient is observed to increase on decreasing the temperature from 299 to 250 K; for 205 mbar and 500 mbar O2 partial pressure, the temperature dependence of the rate coefficient was found to be strongly. Figure 2 shows the temperature dependence of the rate coefficient for OH + DMS obtained in this work in 1000 mbar air and at an oxygen partial pressure of 500 mbar. For comparison this figure includes the fit of Hynes et al. (1986) in 1 atm of air and the values calculated for 205 mbar oxygen partial pressure and 500 mbar oxygen partial pressure using the new preferred expressions from the review of Atkinson et al. (2004). From the Figure 2 it can be seen that at 1 atm of air and in the temperature range from 270 to 299 K, the values obtained in this study are in reasonable agreement with those obtained from the fit of Hynes et al. (1986); but at low temperatures the values measured in this work are higher in comparison with those of Hynes et al. (1986). The values obtained here are, however, in reasonable agreement with the new preferred values from the review of Atkinson et al. (2004), which took into account the data from the new work of Williams et al. (2001). 4 Hynes et al.(1986) fit: 760 Torr of air

Atkinson et al. (2004) (pO2 = 205 mbar)

3

This work (pO2 = 500 mbar)

Atkinson et al. (2004) (pO2 = 500 mbar)

2

kDMS [10

-11

3

-1

-1

cm molecule s ]

This work (1000 mbar of air)

1

0 230

240

250

260

270

280

290

300

310

T (K)

Figure 2.

Variation in kDMS as a function of temperature.

320

228

M. Albu et al.

10

Hynes et al.(1986) fit: 760 Torr of air This w ork (205 mbar O2) + Hynes et al. (1986) - abstraction This w ork (500 mbar O2) + Hynes et al. (1986) - abstraction Williams et al.(2001): 150 Torr of O2 at 600 Torr total pressure This w ork (205 mbar O2) + Atkinson et al. (2004) - abstraction This w ork (500 mbar O2) + Atkinson et al. (2004) - abstraction

branching ratio (addition/abstraction)

8

6

4

2

0 230

240

250

260

270

280

290

300

310

320

T (K)

Figure 3.

Temperature dependence of the branching ratio between the addition and abstraction channel.

Figure 3 shows the temperature dependence of the branching ratio between the addition and abstraction channel for DMS + OH obtained using the data from this work, Hynes et al. (1986), Williams et al. (2001) and Atkinson et al. (2004). For the calculation of the branching ratio between addition and abstraction for this work, values for the abstraction rate coefficients were taken, for comparison purposes, both from the expression reported by Hynes et al. (1986) and also from the new recommendations of Atkinson et al. (2004). The comparison in Figure 3 shows that our data are in good agreement with the more recent data of Williams et al. (2001) and that the work of Hynes et al. (1986) underestimated the ratio between addition and abstraction at low temperatures. Conclusions In this work rate coefficients for the reaction OH + DMS ĺ products have been determined over different conditions of temperature (250 K - 299 K) and oxygen partial pressure (~0 mbar - 500 mbar). For all temperature studied, the rate coefficient increased as the partial pressure of oxygen was increased. The kinetic results in the temperature range (270 K - 299 K) are in good agreement with the work of Hynes et al. (1986). The results at low temperature support the more recent observation of Williams et al. (2001) for a large positive roll-off in the rate coefficient for OH + DMS compared with the work of Hynes et al. (1986).

Kinetic Study of the Temperature Dependence of the OH Initiated Oxidation

229

Acknowledgement Financial support for this work within the EL CID 5th Framework Programme Project of the European Commission is gratefully acknowledged. References Abbatt, J.P.D., F.F. Fenter and J.G. Anderson; High-pressure discharge flow kinetics study of OH + CH3SCH3, CH3SSCH3 o products from 297 to 368 K, J. Phys. Chem. 96 (1992), 1780-1785. Atkinson, R., R.A. Perry and J.N. Pitts Jr.; Rate constants for the reaction of OH radicals with COS, CS2 and CH3SCH3 over the temperature range 299-430 K, Chem. Phys. Lett. 54 (1978), 14-18. Atkinson, R., J.N. Pitts Jr. and S.M. Aschmann; Tropospheric reactions of dimethyl sulfide with NO3 and OH radicals, J. Phys. Chem. 88 (1984), 1584-1587. Atkinson, R.; Gas-phase tropospheric chemistry of volatile organic compounds: 1. alkanes and alkenes, J. Phys. Chem. Ref. Data 26 (1997), 215-290. Atkinson, R., D.L. Baulch, R.A. Cox, J.N. Crowley, R.F. Hampson, R.G. Hynes, M.E. Jenkin, M.J. Rossi and J. Troe; Evaluated kinetic and photochemical data for atmospheric chemistry: volume 1 – gas phase reactions of Ox, HOx, NOx and SOx species, Atmos. Chem. Phys. 4 (2004), 1461-1738. Barnes, I., V. Bastian and K.H. Becker; Kinetics and mechanisms of the reaction of OH radicals with dimethyl sulphide, Int. J. Chem. Kinet. 20 (1988), 415-431. Barone, S.B., A.A. Turnipseed and A.R. Ravishankara; Reaction of OH with dimethyl sulfide (DMS). 1. Equilibrium constant for OH + DMS reaction and the kinetics of the OH.DMS + O2 reaction, J. Phys. Chem. 100 (1996), 14694-14702. Berresheim, H., P.H. Wine and D.D. Davis; Sulphur in the atmosphere, in: H.B. Singh (ed.), Composition, Chemistry, and Climate of the Atmosphere, Van Nostrand Reinhold, New York (1995), 251-307. Charlson, R.J., J.E. Lovelock, M.O. Andreae and S.G. Warren; Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate, Nature 326 (1987), 655-661. Cox, R.A. and D. Sheppard; Reactions of OH radicals with gaseous sulphur compounds, Nature 284 (1980), 330-331. Hsu, Y.-C., D.-S. Chen and Y.-P. Lee; Rate constant for the reaction of OH radicals with dimethyl sulfide, Int. J. Chem. Kinet. 19 (1987), 1073-1082. Hynes, A.J., P.H. Wine and D.H. Semmes; Kinetics and mechanism of OH reactions with organic sulphides, J. Phys. Chem. 90 (1986), 4148-4156. Hynes, A.J., R.B. Stoker, A.J. Pounds, T. McKay, J.D. Bradshaw, J.M. Nicovich and P.H. Wine; A mechanistic study of the reaction of OH with dimethyl-d6 sulfide. Direct observation of adduct formation and the kinetics of the adduct reaction with O2, J. Phys. Chem. 99 (1995), 16967-16975. Kurylo, M.J.; Flash photolysis resonance fluorescence investigation of the reaction of OH radicals with dimethyl sulfide, Chem. Phys. Lett. 58 (1978), 233-237. MacLeod, H., J.L. Jourdain, G. Poulet and G. Le Bras; Kinetic study of reactions of some organic sulfur comounds with OH radicals, Atmos. Environ. 18 (1984), 2621-2626. Martin, D., J.L. Jourdain and G. LeBras; Kinetic study for the reactions of OH radicals with dimethylsulfide, diethylsulfide, tetrahydrothiophene, and thiophene, Int. J. Chem. Kin. 17 (1985), 1247-1261. Nielsen, O.J., H.W. Sidebottom, L. Nelson, J.J. Treacy and D.J. O’Farrell; An absolute and relative rate study of the reaction of OH radicals with dimethyl sulfide, Int. J. Chem. Kin. 21 (1989), 1101-1112. Urbanski, S.P. and P.H. Wine; Chemistry of gas phase organic sulfur-centred radicals, in: Z.B. Alfassi (ed.), SCentred Radicals, John Wiley and Sons (1999), 97-140. Wallington, T.J., R. Atkinson, E.C. Tuazon and S.M. Aschmann; The reaction of OH radicals with dimethyl sulfide, Int. J. Chem. Kin. 18 (1986), 837-846.

230

M. Albu et al.

Williams, M.B., P. Campuzano-Jost, D. Bauer and A.J. Hynes; Kinetic and mechanistic studies of the OHinitiated oxidation of dimethylsulfide at low temperature – a reevaluation of the rate coefficient and branching ratio, Chem. Phys. Lett. 344 (2001), 61-67. Wine, P.H., N.M. Kreutter, C.A. Gump and A.R. Ravishankara; Kinetics of OH reactions with the atmospheric sulfur compounds H2S, CH3SH, CH3SCH3, and CH3SSCH3, J. Phys. Chem. 85 (1981), 2660-2665.

Environmental Chamber Studies of Ozone Formation Potentials of Volatile Organic Compounds William P. L. Carter College of Engineering Center for Environmental Research and Technology, University of California, Riverside, California, 92521, USA Key Words: Environmental chambers, Atmospheric chemical mechanisms, Ozone, Aromatic Hydrocarbons, Petroleum Distillates, Coatings VOCs; SAPRC-99 Mechanism

Abstract Selected results of environmental chamber experiments carried out in the new large indoor environmental chamber at the University of California at Riverside (UCR) that are relevant to quantifying ozone impacts of volatile organic compounds (VOCs) are described. Issues and data needs for quantification of VOC reactivities towards ground-level ozone are described. The ability of the current SAPRC-99 chemical mechanism to simulate recent data from this chamber concerning the ozone formation from irradiations of an ambient reactive organic gas (ROG) - NOx mixture and several new aromatics experiments are discussed. It was found that the mechanism consistently under predicts ozone formation in ambient surrogate - NOx experiments at low ROG/NOx ratios, and also under predicts the effects of adding CO to aromatic - NOx irradiations. The two problems may be related and suggest problems with the formulation of current mechanisms for atmospheric reactions of aromatics. Background Many different types of volatile organic compounds (VOCs) are emitted into the atmosphere, where they can affect photochemical ozone formation and other measures of air quality. Because they can react in the atmospheres at different rates and with different mechanisms, the different types of VOCs can vary significantly in their effects on air quality. The effect of a VOC on ozone formation in a particular environment can be measured by its “incremental reactivity”, which is defined as the amount of additional ozone formed when a small amount of the VOC is added to the environment, divided by the amount added. Although this can be measured in environmental chamber experiments, incremental reactivities in such experiment cannot be assumed to be the same as incremental reactivities in the atmosphere (Carter and Atkinson, 1989; Carter et al., 1995). This is because it is not currently practical to duplicate in an experiment all the environmental factors that affect relative reactivities; and, even if it were, the results would only be applicable to a single type of environment. The only practical means to assess atmospheric reactivity, and how it varies among different environments, is to estimate its atmospheric ozone impacts using airshed models. However, airshed model calculations are no more reliable than the chemical mechanisms upon which they are based. While the initial atmospheric reaction rates for most VOCs are reasonably well known or at least can be estimated, for most VOCs the subsequent reactions of the radicals formed are complex and have uncertainties that can significantly affect predictions of atmospheric impacts. For this reason, environmental chamber experiments and other experimental measurements of reactivity are necessary to test and 231 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 231–240. © 2006 Springer. Printed in the Netherlands.

232

W. P. L. Carter

verify the predictive capabilities of the chemical mechanisms used to calculate atmospheric reactivities. As discussed in a companion paper (Carter, 2004a), under U.S. EPA funding the “UCR EPA” chamber has been constructed at the University of California at Riverside (UCR) in order to address some of the deficiencies of environmental chambers previously used for mechanism evaluation. This chamber has been generating mechanism evaluation data since early 2003, both for general ozone formation mechanism evaluation, and for evaluation of ozone impacts of selected coatings VOCs. The data have been used to evaluate the SAPRC-99 gas-phase mechanism (Carter, 2000), which, together with the European Master Chemical Mechanism (MCM) (Jenkin et al, 1997; Saunders et al, 2003; Jenkin et al, 2003; MCM 2004) represents one of the two state-of-the-art mechanisms that can be used to assess ozone impacts of many individual VOCs. Although the evaluation of the MCM against chamber data is currently underway (e.g., Pilling, 2004; Pinho et al, 2003), SAPRC-99 is still the most extensively evaluated of the mechanisms against available chamber data (Carter, 2000). In this paper, we will discuss the results of the evaluations of the SAPRC-99 mechanism with UCR EPA chamber data relevant to chemical mechanism uncertainty with respect to predictions of effects of VOCs on ozone. The data discussed here are relevant “base mechanism” uncertainty, which refer to general uncertainties in the predictions of ozone formation from reactive organic gas (ROG) and NOx mixtures representing ambient pollution. This is important to prediction of effects of overall VOC and NOx controls and also to appropriately representing the chemistry of base case conditions when assessing incremental reactivities of various types of individual types of VOCs. The specific experiments discussed here are ambient ROG surrogate - NOx experiments at varying ROG and NOx levels, and experiments to evaluate the photooxidation mechanism for aromatics, which are important and highly reactive components of the ambient ROG mixture. Prediction of Ozone formation in Ambient ROG Surrogate - NOx Experiments As discussed in our companion presentation (Carter, 2004a), the initial experiments carried out in the UCR EPA chamber consisted of a large number of ambient ROG surrogate NOx experiments carried out at varying initial NOx and ROG levels. The ambient ROG surrogate composition was derived as discussed by Carter et al (1995) and consisted of a simplified mixture of designed to represent the major classes of hydrocarbons and aldehydes measured in ambient urban atmospheres, with one compound used to represent each model species used in current condensed lumped-molecule mechanisms. The eight representative compounds used were n-butane, n-octane, ethene, propene, trans-2-butene, toluene, m-xylene, and formaldehyde. The initial NOx levels varied from 2 to ~300 ppb and the initial ROG levels varied from ~0.2 to 4.2 ppmC. Figure 1 shows the matrix of initial ROG and NOx levels of the ambient ROG surrogate - NOx experiments carried out in the UCR EPA chamber for mechanism evaluation. The different symbols indicate the project for which the experiment was conducted and whether radical measurement data were made, but for the purpose of this discussion they all are the same experiment except for the varying ROG and NOx levels. Note that this is a loglog plot, with the initial NOx varying by almost two orders of magnitude, and the initial ROG varying by over one order of magnitude. The solid line show the ROG/NOx levels that the SAPRC-99 mechanism predicts give the highest ozone levels and the dotted line shows the

Environmental Chamber Studies of Ozone Formation Potentials

233

ROG/NOx levels where the mechanism predicts O3 has the greatest sensitive to changes in VOC levels (MIR line). It can be seen that the matrix includes experiments in both the VOC sensitive and NOx-sensitive regions as defined by these lines.

Initial NOx (ppb)

1000

Initial Evaluation Runs

VOC Sensitive Region

OBM Runs without Radical Data

100

OBM Runs with OH, HO2 Data Used for Base Case for Reactivity Runs

10

Calculated Maximum O3 Line

NOx-Limited Region

Calculated MIR Line

1 0.1

1

10

Initial ROG (ppmC)

Figure 1.

Matrix of initial ROG and NOx levels in the ambient ROG surrogate - NOx experiments carried out in the UCR EPA chamber. The initial ROG and NOx levels used in the base case experiments for the incremental reactivity studies of the coatings VOC emissions are also shown.

The two large open circles on Figure 1 show the ROG and NOx levels chosen to serve as the “base case” in the incremental reactivity experiments that are currently underway in this chamber to assess ozone impacts of various different types of VOCs (Carter, 2004c). The 2530 ppb NOx levels were chosen to be representative of pollution episodes of interest in California, based on input provided by the staff of the California Air Resources Board, which is funding most of the current reactivity studies. The ROG/NOx ratios were chosen to represent two sets of conditions of NOx availability relevant to VOC reactivity. The lower ROG/NOx ratio was chosen to represent the relatively higher NOx conditions of maximum incremental reactivity (MIR) where ozone is most sensitive to VOCs, to approximate the conditions used to derive the widely-used MIR ozone reactivity scale (Carter, 1994). The higher ROG/NOx ratio was chosen to be one-half that yielding maximum ozone levels, and is used to represent conditions where ozone is NOx-limited, but not so NOx-limited that VOC reactivity is irrelevant. These are referred to in the subsequent discussion as the “MIR” and “MOIR/2” base cases, respectively. The ability of the SAPRC-99 mechanism to simulate ozone formation in representative MIR and MOIR/2 base case experiments is shown on Figure 2. It can be seen that good reproducibility in ozone formation is observed in the replicate experiments, but that the model consistently under predicts O3 formation in the MIR experiments, though it gives much better predictions of ozone in the experiment at the higher ROG/NOx ratio. This tendency to under

234

W. P. L. Carter

MIR Base Case

MOIR/2 Base Case

(NOx = 30 ppb, ROG = 0.5 ppmC)

(NOx = 25 ppb, ROG = 1 ppmC)

0.16

0.16

0.12

0.12

Model

0.08

EPA110B

EPA114A

0.04

Ozone (ppm)

Ozone (ppm)

predict O3 in the base case experiment representing MIR conditions presents a problem when using incremental reactivity MIR experiments for evaluating mechanism of added VOCs.

Model

0.08

EPA095B

0.04

EPA124B

EPA123A

EPA127B

EPA126A 0.00

0.00 0

60

120

180

240

Time (minutes)

Figure 2.

300

360

0

60

120

180

240

300

360

Time (minutes)

Experimental and calculated concentration-time plots for ozone representative base case experiments used for VOC reactivity assessment.

in

The dependence of the tendency for the SAPRC-99 mechanism under prediction bias on the initial ROG/NOx ratio is shown on Figure 3. In order to provide useful mechanism evaluations at the lowest ROG/NOx ratios where O3 formation is suppressed by NO, the model predictions of ozone formed + NO oxidized, measured by ([O3]final-[NO]final) ([O3]initial-[NO]initial) or '([O3]-[NO]), is shown. Comparable results of a recent SAPRC-99 mechanism evaluation using ambient surrogate - NOx data from other low NOx environmental chamber datasets (Carter, 2004b, and references therein) are also shown. In order to place the ROG/NOx ratios for the various chamber datasets on the same basis, the biases are plotted against ROG/NOx ratios normalized to the ratio calculated to give the highest ozone formations for the conditions of the various chambers. Although this was not evident from the previous mechanism evaluation using higher NOx chamber data with higher chamber effects (Carter, 2000; see also Carter, 2004b), the results of this work show a consistent tendency of increase in ozone under prediction bias with decreasing ROG/NOx ratios for the ambient surrogate experiments in the UCR EPA chamber. This bias may also be present in the simulations of the ambient surrogate experiments in the TVA chamber, which were also carried out at relatively low NOx levels (Simonaitis and Bailey, 1995; Bailey et al, 1996; Carter, 2004b). The SAPRC-99 mechanism under predicts O3 formation and NO oxidation in almost all the ambient surrogate experiments carried out in the Australian CSIRO chamber that were used for low NOx mechanism evaluation (Hess et al, 1992; Johnson et al, 1997; Carter, 2004b). Although this might be due to characterization uncertainties with this outdoor chamber (Carter, 2004b), if the relative ROG/NOx ratios are taken into account the results are entirely consistent with the trends observed when modelling the UCR EPA experiments.

Environmental Chamber Studies of Ozone Formation Potentials

235

Plots of the bias for the SAPRC-99 mechanism to under predict NO oxidation and O3 formation against normalized ROG/NOx ratios for ambient surrogate NOx experiments in three environmental chambers.

'(O3-NO) Model underprediction Bias

Figure 3.

80%

SAPRC-99 (Range shows effect of varying HONO offgasing rates)

MOIR

60% SAPRC-99 with Reduced k(OH + NO2)

40%

Carbon Bond 4 Mechanism

20% 0% Approx. MIR

-20% 0.1

Figure 4.

1

10

Plots of the mechanism under prediction bias for model simulations UCR EPA ambient surrogate - NOx experiments against normalized ROG/NOx ratios for the SAPRC-99 and Carbon Bond mechanisms. The SAPRC-99 calculations show effects of varying the HONO offgasing chamber effect parameters and using a lower OH + NO2 rate constant.

Figure 4 shows the effects of using different mechanisms or mechanism parameters on plots of the under prediction bias the UCR EPA ambient surrogate experiments. The filled diamonds show the same results as shown on Figure 3 for this chamber, and the error bars show the effects of varying the HONO offgasing chamber effect parameter from zero to the

236

W. P. L. Carter

maximum level indicated by the variability of the characterization experiments. This is probably the most important chamber effects parameter that would affect the results of model simulations of these experiments, since the radical source caused by the HONO photolysis will affect the results of the low ROG/NOx experiments while the NOx input caused by this process will affect the high ROG/NOx runs. While this variable chamber effect has a nonnegligible effect, it is not enough to explain the under prediction bias observed at the low ROG/NOx ratios. On the other hand, it could account for the apparent over prediction bias at the highest ROG/NOx ratios. Various sensitivity studies have indicated that uncertainties in the rate constant for the reaction of OH with NO2, an important radical termination and NOx process, could have effects of model simulations of O3. In order to assess whether this could account for the bias observed, simulations were carried out with the rate constant reduced by ~20%, which is probably reflective of its current uncertainty. It can be seen that this reduces the bias somewhat, but does not eliminate it. Figure 4 also shows that the low ROG/NOx model under prediction bias is significantly worse for the Carbon Bond 4 mechanism (Gery et al, 1989). Although somewhat out-of-date, this mechanism is still widely used in many models in the United States because of its computational efficiency. The reason for this under prediction bias is still uncertain, and it is being investigated by conducting experiments with modified ROG surrogates and as part of an ongoing project to update the SAPRC mechanism. We suspect that is probably due to problems with the current mechanism for aromatics, which are highly reactive components of the ambient ROG surrogate whose mechanism remain highly uncertain. Additional evidence for problems with the current aromatics mechanisms is discussed in the following section. New Aromatics Mechanism Evaluation Results As discussed in our companion paper (Carter, 2004a), the initial evaluation experiments in the UCR EPA chamber included experiments with the representative aromatic hydrocarbons toluene and m-xylene. Aromatics are important components of ambient ROG mixtures because of their high reactivity, and both toluene and m-xylene are present in the ambient ROG surrogate used in the experiments discussed in the previous section. Aromatics have the most uncertain mechanisms of all the compounds present in this surrogate. Although there have been extensive studies of the atmospheric reactions of aromatics and their reactive products (e.g., see Calvert et al, 2002, and references therein), available information is still insufficient to derive predictive mechanisms. The greatest uncertainty concerns the highly reactive but incompletely characterized aromatic ring-opening products, whose subsequent reactions appear to be significant radical initiators, since ignoring their reactions results in significant under predictions of aromatic reactivity. The current SAPRC-99 mechanism and represents the reactions in a highly simplified manner, using a few parameterized model species whose yields and photolysis rates were adjusted to optimize fits of model simulations to aromatic - NOx chamber experiments (Carter, 2000). The results of the toluene - NOx and m-xylene - NOx experiments in the UCR EPA were reasonably consistent with the predictions of the current SAPRC-99 mechanism, though the mechanism had a slight bias towards over predicting O3 in the experiments (see Carter, 2004a for a summary of the fits to total NO oxidation and O3 formation in the experiments).

Environmental Chamber Studies of Ozone Formation Potentials

237

Examples of experimental and calculated ozone in these runs are shown in Figure 5, where the open squares are the experimental data for representative aromatics - NOx experiments, and the dotted lines are the model simulations of those runs. A reasonably good fit of the model to these experiments is expected because the parameters in the mechanisms for those compounds were adjusted to optimize model simulations of such experiments. However, the NOx levels in these experiments were considerably lower than those in the runs used to develop the SAPRC-99 mechanisms, and there was a concern that the parameterization may not extrapolate to lower NOx levels. Apparently the parameterization employed is no less applicable for NOx levels as low as ~5 ppb than it is for the higher NOx levels in the experiments used in the mechanism development. Toluene

m-Xylene

Concentration (ppm)

Ozone Change Ozone (effect of CO) EPA077 (~25 ppb NOx)

Ozone Change (effect of CO) EPA067 (~20 ppb NOx)

Ozone

0.12

0.25

0.06

0.12 0.20

0.09 0.04

0.08

0.15

0.06 0.10

0.02

0.04

0.03

0.05

0.00

0.00 0

120

240

0.00

0

360

120

240

360

Time (minutes)

0.00

0

120

240

360

0

120

240

360

Time (minutes)

Concentration (ppm)

EPA066 (~5 ppb NOx) 0.10

0.04

0.08

0.03

0.06 0.04

Aromatic - NOx experiment

0.02

Aromatic - CO - NOx experiment

0.01

Aromatic - NOx model

0.02

0.00 0.00

0 0

180

360

540

180

360

540

Aromatic - CO - NOx model

-0.01

Time (minutes)

Figure 5.

Experimental and calculated effects of CO addition on selected aromatics NOx experiments carried out in the UCR EPA chamber.

However, the new UCR EPA chamber dataset also includes experiments where the effect of adding CO to aromatics - NOx irradiations was determined. These experiments were carried out because model calculations indicated that the addition of CO would cause a significant enhancement in O3 formation, due to the NO to NO2 conversions caused when CO reacts with the radicals produced in the aromatic photooxidation reactions, and we wanted to test this prediction. In this regard, CO acts as a “radical amplifier”, enhancing the effects of radicals on ozone formation. CO addition is also useful because CO has the simplest possible mechanism to represent other VOCs present in ambient mixtures; its reactions only cause NO to NO2 conversions and its reactions result in formation of no other products or direct radical sources or sinks. Therefore, added CO experiments should provide a test of an aspect of the aromatics mechanisms that is applicable to its effects in ambient simulations, and that is different than the tests provided by aromatic NOx experiments in the absence of other VOCs.

238

W. P. L. Carter

The results of the aromatics - NOx experiments with the added CO were consistent with the model in that the added CO did indeed significantly increase ozone formation, but the amount of O3 increase was about twice what the model predicted. This is shown on Figure 5, where the filled diamonds show the O3 in the aromatic - NOx experiments or the change in O3 caused by adding the CO, and the solid lines show the corresponding model predictions. About the same under prediction of the CO addition effect is seen for m-xylene as toluene, but the under prediction may be greater in the lower NOx toluene experiment. The under prediction of the effect of the added CO suggests that the mechanism is under predicting the radical input in the aromatic photooxidation process. The fact that the same mechanism can simulate O3 formation in the aromatics - NOx experiments without the CO indicates that there probably are compensating errors in the mechanism. Further evidence for this comes from the fact that the mechanism overpredicts NO to NO2 conversions in “direct reactivity” aromatic + HONO experiments that are much more sensitive to NO to NO2 conversions in the mechanisms than radical sources (Carter and Malkina, 2002), suggesting that there may be too many NO to NO2 conversions in the aromatics mechanisms. However, reducing the NO to NO2 conversions in the SAPRC-99 aromatics mechanism to fit the direct reactivity data and re-adjusting the radical input parameters to simulate the aromatics - NOx experiments does not fully correct this problem. This is shown on Figure 6, which shows the model predictions of effects of CO, direct reactivity, and O3 formation in toluene - NOx runs with the standard and modified SAPRC-99 toluene mechanisms. Although the modified mechanism predicts an increase in O3 caused by the added CO compared to the standard mechanism, the increase is insufficient. Even increasing the radical input in the model to levels that result in over prediction of O3 in the toluene - NOx experiments is still not sufficient for the model to simulate the full effect of the CO addition. Effect of CO

Direct Reactivity

Toluene - NOx Runs

(EPA074B vs EPA074A)

0.05

0.00 0

120 240 Time (minutes)

360

0.3 0.4

0.2 0.1

0.2

0.0 2

4 6 Toluene Added (ppm)

Data

Standard Mechanism Re-Adjusted

8

'([O3]-[NO]) (ppm)

0.10

CTC079

0.4

0.6

'([O3]-[NO]) Change (ppm)

'([O3]-[NO]) Change (ppm)

0.15

0.0 0

120

240

360

CTC026

0.6 0.4 0.2

Radical Yield Increased Further 0.0 0

120

240

360

Time (minutes)

Figure 6.

Effects of adjustments to the SAPRC-99 toluene photooxidation mechanism on simulations of representative toluene experiments experiments.

Environmental Chamber Studies of Ozone Formation Potentials

239

Thus the new data suggest that there is a fundamental problem with the formulation of the current aromatics mechanism, and more extensive modifications are required in order for model simulations to be consistent with the available data. We have been funded by the California Air Resources Board to update the SAPRC mechanism, and developing an improved aromatics mechanism is a priority in this project. Work on a new, more explicit mechanism that incorporates more recent laboratory data such as that summarized by Calvert et al (2004) is now underway, but progress is slow. Presently, we are encountering problems similar to those discussed by Pilling (2004) for the MCM. It is likely that there are reactions occurring that none of the mechanisms are adequately representing. Conclusions Although the UCR EPA chamber has only been in operation for a relatively short time, it has already obtained useful information concerning the performance of current mechanism in predicting the effects of VOCs and NOx on ozone formation. As discussed here and also in our companion paper (Carter, 2004a), the SAPRC-99 mechanism predicts O3 formation reasonably well in low NOx experiments, and in ambient simulation experiments at the high ROG/NOx levels where maximum ozone formation potentials are achieved. However, the new data indicate problems with the mechanisms that were not previously realized. The SAPRC99 mechanism consistently under predicts O3 formation in the lower ROG/NOx experiments where O3 formation is most sensitive to VOCs, and the problem is even worse for CB4. Other experiments indicate that there are problems with the formulation with the current aromatics photooxidation mechanisms. It is possible that the problems with the under prediction at low ROG/NOx ratios may be caused by problems with the aromatics mechanisms. Experimental and mechanism development work to investigate and hopefully resolve these problems is underway References Bailey, E. M., C. H. Copeland and R. Simonaitis; Smog chamber studies at low VOC and NOx concentrations,” Report on Interagency Agreement DW64936024 to EPA/NREL, Research Triangle Park, NC. (1996). Calvert, J. G., R. Atkinson, K. K. Becker, R. M. Kamens, J. H. Seinfeld, T. J. Wallington, and G. Yarwood; The mechanisms of atmospheric oxidation of aromatic hydrocarbons, New York: Oxford University Press, 556 pgs. (2002) Carter, W. P. L.; Development of ozone reactivity scales for volatile organic compounds,” J. Air & Waste Manage. Assoc., 44, 881-899. (1994) Carter, W. P. L.; Documentation of the SAPRC-99 chemical mechanism for VOC reactivity assessment,” Report to the California Air Resources Board, Contracts 92-329 and 95-308, May 8 (2000). Available at http://www.cert.ucr.edu/~carter/absts.htm#saprc99. Carter, W. P. L.; The UCR EPA Environmental Chamber. Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004a). Carter, W. P. L.; Evaluation of a gas-phase atmospheric reaction mechanism for low NOx conditions,” Final Report to California Air Resources Board Contract No. 01-305, May 5 (2004b). Available at http://www.cert.ucr.edu/~carter/absts.htm#lnoxrptf Carter, W. P. L.: Main information page for “Evaluation of atmospheric ozone impacts of selected coatings VOC emissions” and “Environmental chamber studies of VOC species in architectural coatings and mobile source emissions”. http://www.cert.ucr.edu/~carter/coatings. Last modified April 17. (2004c)

240

W. P. L. Carter

Carter, W. P. L. and R. Atkinson; A computer modeling study of incremental hydrocarbon reactivity, Environ. Sci. Technol., 23, 864 (1989) Carter, W. P. L., D. Luo, I. L. Malkina, and J. A. Pierce; Environmental chamber studies of atmospheric reactivities of volatile organic compounds. effects of varying ROG surrogate and NOx,” Final report to Coordinating Research Council, Inc., Project ME-9, California Air Resources Board, Contract A0320692, and South Coast Air Quality Management District, Contract C91323. March 24 (1995). Available at http://www.cert.ucr.edu/~carter/absts.htm#rct2rept. Carter, W. P. L. and I. L. Malkina; Development and application of improved methods for measurement of ozone formation potentials of volatile organic compounds, Final report to California Air Resources Board Contract 97-314, May 22 (2002). Available at http://www.cert.ucr.edu/~carter/absts.htm #rmethrpt. Hess G. D., F. Carnovale. M. E., Cope. and G. M. Johnson; The evaluation of some photochemical smog reaction mechanisms – 1 temperature and initial composition effects, Atmos. Environ., 26A, 625-641. (1992) Jenkin, M. E., S. M. Saunders and M. J. Pilling; The tropospheric degradation of volatile organic compounds : a protocol for mechanism development,” Atmos. Environ. 31, 81-104 (1997). Jenkin, M.E., S.M. Saunders, V. Wagner and M.J. Pilling; Protocol for the development of the master chemical mechanism MCMv3 (Part B): Tropospheric degradation of aromatic volatile organic compounds, Atmospheric Chemistry and Physics Discussions, 2, p1905-1938 (2002). Atmospheric Chemistry and Physics, 3, 181-193 (2003) Johnson G., P. C. Nancarrow, A.Quintanar, and M. Azzi; Smog chamber data for testing chemical mechanisms at low VOC to NOX ratio conditions, Final Report to EPA Cooperative Agreement CR-821420, Atmospheric Research and Exposure Assessment Laboratory, U.S. EPA. (1997) MCM; The Master Chemical Mechanism website, http://chmlin9.leeds.ac.uk/MCM/. Undated. Last accessed 10/2004 (2004) Pihno, P. G.; C. A. Pio, and M. E. Jenkin; Evaluation of isoprene degradation on the detailed tropospheric chemical mechanism, MCM v3, using environmental chamber data. Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004). Pilling, M. J.; Aromatic hydrocarbon oxidation: the contribution of chamber photooxidation studies. Presented at the NATO EST-ARW Workshop on Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, Zakopane, Poland, October 1-4 (2004). Saunders, S.M., M.E. Jenkin, R.G. Derwent and M.J. Pilling; Protocol for the development of the master chemical mechanism MCMv3 (Part A): Tropospheric degradation of non-aromatic volatile organic compounds, Atmospheric Chemistry and Physics Discussions, 2, p1847-1903 (2002). Atmospheric Chemistry and Physics, 3, 161-180 (2003) Simonaitis, R. and E. M. Bailey; Smog chamber studies at low VOC and NOx concentrations: Phase I,” Report on Interagency Agreement DW64936024 to EPA/NREL, Research Triangle Park, NC. (1995).

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3, Using Environmental Chamber Data: Butane and Its Degradation Products P. G. Pinho1, C.A. Pio1, and M. E. Jenkin2 2

1 Departamento de Ambiente, Escola Superior de Tecnologia de Viseu, 3504-510 Viseu, Portugal Department of Environmental Science and Technology, Imperial Collage London, Silwood Park, Ascot, Berkshire SL5 7PY, UK

Key Words: Butane, Environmental Chamber, Mechanisms, Modelling, Ozone, Troposphere

Introduction The chemistry of ground level atmospheric ozone formation is highly complex and nonlinear and has many uncertainties. As a result no chemical model can be relied upon to give accurate predictions unless it has been evaluated by comparison with experimental data. There are essentially two ways in which a photochemical oxidant model can be evaluated. One compares the predictions of the complete model against field data taken during a historic ozone pollution episode. Another approach relies on the evaluation of individual pollutant compounds, separately. In the case of gas-phase chemical mechanisms, this means evaluating the predictions of the mechanism against results of environmental chamber experiments (Carter et al., 1995a). There are a large number of condensed chemical reaction mechanisms to describe and represent atmospheric chemical transformations involving air pollutants. On the contrary only a very limited number of mechanisms exist with a detailed description of chemical transformation processes for primary and secondary atmospheric species. The Master Chemical Mechanism (MCM) (Jenkin et al., 1997) and SAPRC (Carter and Lurmann, 1991) are two well known examples. The MCM is a near-explicit chemical mechanism that describes the detailed degradation of a series of emitted VOC, and the resultant generation of ozone and other secondary pollutants, under conditions appropriate to the planetary boundary layer. Version 3 of the Master Chemical Mechanism (MCM v3) considers the oxidation of 125 VOC. The complete mechanism comprises 12,691 reactions of 4,351 organic species and 46 associated inorganic reactions, which were defined on the basis of the MCM scheme writing protocols (Jenkin et al., 1997, Saunders et al., 2003; Jenkin et al., 2003). Although MCM v3 has recently been superseded by v 3.1, the chemistry for non-aromatic VOC remains unchanged. Although the entire MCM has been tested in atmospheric models, and through intercomparison with the results of chamber-validated mechanisms (e.g. Derwent et al., 1998; Jenkin et al., 2002), it has only been partially tested using environmental chamber data. It has been used in a number of studies involving the European Photoreactor (EUPHORE) in Valencia (EUPHORE, 2002), providing the basis for validation of the mechanisms for selected VOC. MCM v3 chemistry has thus already been tested for the photo-oxidation of Dpinene-NOx mixtures at comparatively low NOx concentrations (Saunders et al., 2003). The aromatic mechanisms in MCM v3 and MCM v3.1 have also been evaluated against a set of smog-chamber experiments; the evaluation was focused on four representative species of the 241 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 241–252. © 2006 Springer. Printed in the Netherlands.

P. G. Pinho et al.

242

class of aromatic compounds: benzene, toluene, p-xylene and 1,3,5-TMB (Bloss et al., 2004a; 2004b). The main objective of the present study is the evaluation and refinement of MCM v3, using the large environmental chamber dataset of the Statewide Air Pollution Research Center (SAPRC) at the University of California (Carter et al., 1995a). Initial evaluation of the representation of chamber-dependent processes has been carried out using butane-NOX photooxidation data and the MCM v3 degradation chemistry for butane, and the results of this evaluation are presented. Within this assessment, photooxidation experiments involving the major carbonyl products formed during butane degradation, namely methylethyl ketone (MEK), acetaldehyde (CH3CHO) and formaldehyde (HCHO) were also considered. Throughout the study, the performance of the MCM v3 chemistry was simultaneously compared with that of the corresponding chemistry in the SAPRC-99 mechanism, which was developed and optimized in conjunction with the chamber datasets (Carter, 2000). This work therefore develops a base for the evaluation of degradation mechanism for a number of VOC represented in MCM v3, using the SAPRC environmental chamber datasets. Chamber data The mechanistic evaluation was carried out using the database of the indoor chambers of the Statewide Air Pollution Research Center (SAPRC) at the University of California (Carter et al., 1995a). A set of computer data files containing the SAPRC environmental chamber data was obtained on the Internet by anonymous FTP at carterpc.ucr.edu. Data from the following environmental chambers were used in the present evaluation: Indoor Teflon Chamber #1 (ITC); Indoor Teflon Chamber #2 (ETC); Dividable Teflon Chamber (DTC); Xenon Arc Teflon Chamber (XTC) and CE-CERT Xenon Arc Teflon Chamber (CTC). The available datasets from these chambers were used previously in the development and evaluation of the SAPRC-99 mechanism (Carter, 2000). A large set of chamber experimental runs were employed in the present evaluation. These consisted of 46 butane-NOX-air, 6 MEK-NOX-air, 11 CH3CHO-NOX-air and 24 HCHO-NOX-air experiments, in the initial evaluation of the chamber dependent processes and butane degradation chemistry. The ranges of reagent concentrations for the considered experiments are presented in Table 1. Full details of the experimental conditions and reagent concentrations are available elsewhere (Carter et al., 1995a, Carter, 2000), which includes detailed documentation of the measurement uncertainties associated with the analytical methods employed for the reactants and products, and the precision and accuracy of other experimental parameters (e.g. temperature; light intensity and spectrum of the photolysing radiation). Chamber-dependent processes Wall effects Heterogeneous processes occurring at the chamber walls are probably the most serious problem in evaluation and interpretation of experiments carried out in environmental chambers. These wall effects are known to be non-negligible in all current-generation chambers, and can dominate the results of certain types of experiments (e.g., Carter and Lurmann, 1991, Carter, 2000).

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3 Table 1.

243

Ranges of reagent concentrations for the considered experiments. Butane-NOx

MEK-NOx

CH3CHO-NOx

HCHO-NOx

[VOC] (ppm)

1.47–4.95

7.83–9.49

0.44–1.67

0.29–1.01

[NOx] (ppm)

0.09-0.66

0.09–0.29

0.14–0.28

0.16-0.54

[VOC]/[NOx]

5.22-39.89

1.71–11.73

1.00–3.40

26.61– 99.44

There are three major groups of processes that contribute to chamber wall effects: the introduction of free radicals through wall reactions, the adsorption or desorption of oxidised nitrogen species (NOy), and off-gassing of organics that can lead to OH to HO2 conversion and therefore supplement ozone production in the system (Killus and Whitten, 1990). The chamber wall effects can be characterized for a given chamber, through experiments which are designed to be especially sensitive to the wall effects, but for which the gas phase chemistry is reasonably well understood. Such characterization experiments can be used to create a model for chamber wall effects and chamber-dependent parameters (Carter et al., 1995a), which is usually termed the ‘auxiliary mechanism’ (Jeffries et al., 1992). A database for mechanism evaluation is therefore not complete without recommendations for appropriate inputs to auxiliary mechanisms. Such chamber effects models have already been developed for the SAPRC chambers as part of the process of evaluating SAPRC-99 and related mechanisms (e.g. Carter, 2000), and these form the basis of the auxiliary mechanisms applied in the present study. Photolysis processes The rates of photolysis processes depend on the intensity and spectral distribution of the light source, which vary from one chamber to another. Spectral distributions (based on spectrometer measurements) and absolute light intensities (based on NO2 actinometry experiments) are documented for all the SAPRC chambers (e.g. Carter et al., 1995a, Carter et al., 1995b). The rates of photolysis processes for application in MCM v3 were thus determined using this spectral information in conjunction with absorption cross sections and quantum yield data previously used in conjunction with the MCM, based on the sources summarised by Jenkin et al., (1997). Furthermore, the rates of a number of photolysis reactions were updated during the course of the present study, as discussed further below. Butane photo-oxidation and auxiliary mechanism assessment Chemistry of butane photo-oxidation The complete degradation chemistry of butane, as represented in MCM v3, consists of 538 reactions of 178 species. It can be viewed and downloaded using the subset mechanism assembling facility, available as part of the MCM website. The methodology of mechanism construction has been described in detail by Jenkin et al., 1997 and Saunders et al., 2003. The main features of the degradation chemistry when NOX is present are summarized in Figure 1.

P. G. Pinho et al.

244

butane OH O2

C2 H 5 O

NO

NO

O2

O

O2

HO2

HO2

O2

O

CH3CHO O

OH

NO

O2

O2

OH

MEK

O

NO2

CH3C(O)O2

- NO2

NO

O2

CH3C(O)O2NO2

PAN

CH3C(O)O O2 NO

CH3O

O2

HO2

OH

HCHO

O2

CO2

O2

OH

CO

HO2

HO2

Figure 1.

Schematic representation of the major radical propagation pathways of the OHinitiated degradation of butane, also illustrating the sequential formation of the intermediate products MEK, CH3CHO, HCHO and CO. Reaction steps involving NO are reactions with organic peroxy radical intermediates, which lead to NO-to-NO2 conversion and O3 production (see discussion in text). The reactions of OH with butane and MEK possess additional channels which are fully represented in the mechanism, but not shown in the diagram.

Mechanism evaluation As with the chamber characterisation studies of Carter (Carter, 2000), the series of butane-NOX photo-oxidation experiments was used for initial assessment of the auxiliary mechanism parameters. This system is believed to provide a good test for the chamber wall effects because the degradation chemistry of butane is quite well characterized (Carter and Lurmann, 1991, Carter et al., 1995a), and because of the large set of experiments for which data are available. Initial simulations were carried out to compare the SAPRC-99 and MCM v3 butane mechanisms for the conditions of the butane-NOX experiments, using the auxiliary mechanism parameters defined by Carter, (2000), and a consistent set of inorganic reaction

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3

245

rate coefficients. The MCM v3 and SAPRC-99 butane mechanisms were found to perform very similarly, suggesting a good level of consistency in the understanding and representation of butane oxidation in these mechanisms. To allow a full evaluation, however, it is necessary to consider not only the experimental data for the butane-NOx system, but also the NOx photo-oxidation runs for intermediate products, MEK, CH3CHO and HCHO. The aim of the mechanism refinements for butane degradation, therefore, was to provide an adequate, self-consistent description of all these systems, whilst simultaneously identifying any required modifications to the auxiliary mechanism parameters. Mechanism testing and refinement was therefore carried out by an iterative procedure, considering butane photo-oxidation and each of the sub-systems in turn. The adapted version of MCM v3, resulting from identified modifications, is denoted ‘MCM v3a’ throughout this article. In conjunction with these tests, the opportunity was taken to revise and update the inorganic chemistry parameters. The most notable change from the reference simulations, as performed for SAPRC-99 by Carter, (2000), was a change in the assigned rate coefficient for the reaction of OH with NO2 from a value (at 298 K and 760 Torr) of 8.98 × 10-12 cm3 molecule-1 s-1 (based on JPL, 1997) to a value of 1.05 × 10-11 cm3 molecule-1 s-1 (based on JPL, 2003). This reaction represents an important radical sink in the system, which was found to have a particularly significant impact under the conditions of some of the butane-NOx photo-oxidation experiments. As described further below, this required small compensating increases to be applied to the chamber radical source parameter. As with previous assessments (e.g. Carter, 2000), the quantity D(O3-NO) was used as the main criterion of model performance. This quantity is defined as: D(O3-NO)t = [O3]t [NO]t - ([O3]0 - [NO]0), where [O3]0, [NO]0, and [O3]t, [NO]t are the concentrations of O3 and NO at the beginning of the run, and at time ‘t’, respectively. As described in detail previously (e.g., Carter and Lurmann, 1991, Carter et al., 1995a, Carter, 2000), D(O3-NO) is an indicator of the ability of the mechanism to simulate the chemical processes that cause O3 formation, giving a useful measure, even when O3 is suppressed by the presence of excess NO. In addition, use of this measure allows a direct comparison with the SAPRC-99 published results. The precursor decay rate and formation of the carbonyl products and PAN were also used as criteria of model performance. HCHO-NOX experiments The MCM v3 mechanism was found to under-predict D(O3-NO) in many (but not all) of the runs, principally for the CTC and DTC chambers. This discrepancy was improved through the implementation of revised absorption cross section and quantum yield data for the photolysis of HCHO. The photolysis parameters were updated in line with the latest IUPAC recommendations (Data Sheet P1 updated in 16th May 2002), the preferred cross sections being based on the data of Meller and Moortgat, (2000), and the quantum yields based on data from Smith et al. (2002). The new cross sections are 5-10% higher than the values previously recommended by IUPAC (and adopted for the MCM by Jenkin et al., 1997). The improved performance of the modified mechanism, MCM v3a, in simulating the HCHO-NOX chamber experiments is summarised in Figure 2 and Figure 3. The improvement is particularly apparent for the large set of CTC and DTC chambers runs, which represent the majority of the dataset for HCHO.

246

P. G. Pinho et al.

0,45

0,30

DTC149B

0,60

CTC024

CTC016

0,40

0,25

0,50

0,20

0,40

0,15

0,30

0,10

0,20

0,05

0,10

0,35 0,30 0,25 0,20 0,15 0,10 0,05

0,00

0,00 0

60

120

180

240

300

0,00

0

360

SAPRC 99

60

120

180

240

300

360

0

60

120

180

240

300

360

MCM v3a

Experimental

MCM v3

Plots of experimental and calculated D(O3-NO) (ppm) vs. time (min) for the HCHO – NOx -air experiments.

Figure 2. 0,50

0,40

DTC149B

0,45

0,90

CTC024

0,35

0,80

0,30

0,70

0,40 0,35 0,30

0,25

0,25

0,20

0,20

0,15

CTC016

0,60 0,50 0,40 0,30

0,15

0,10

0,20

0,10 0,05

0,05

0,00

0,00 0

60

120

180

240

300

0,10 0,00

0

360

60

120

180

240

MCM v3

SAPRC 99

300

360

0

60

120

Experimental

180

240

300

360

MCM v3a

Plots of experimental and calculated HCHO (ppm) vs. time (min) for the HCHO – NOx -air experiments.

Figure 3.

0,25

0,40

DTC387B

0,40

CTC014

0,35

0,35

0,30

0,30

0,25

0,25

0,20

0,20

0,15

0,15

0,10

0,10

0,05

0,05

DTC150B

0,20

0,15

0,10

0,05

0,00

0,00

0,00 0

60

120

180

240

SAPRC 99

Figure 4.

300

360

0

60

MCM v3

120

180

240

300

360

Experimental

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated D(O3-NO) (ppm) vs. time (min) for the CH3CHO – NOx –air experiments.

CH3CHO-NOX experiments Apart from a small secondary influence of the changes in the HCHO photolysis parameters, no further modifications were required to the acetaldehyde mechanism. As shown in Figure 4 to Figure 7, MCM v3a provides a very good description of D(O3-NO) in the complete set of CH3CHO-NOX chamber experiments.

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3 0,6

1,0

DTC387B

0,5

1,8

CTC014

0,9

1,6

0,8

1,4

0,7

0,4

247 DTC150B

1,2

0,6

1,0

0,5

0,3

0,8

0,4 0,2

0,6

0,3

0,4

0,2

0,1

0,2

0,1 0,0

0,0 0

60

120

180

240

300

0,0

0

360

SAPRC 99

60

120

180

240

300

360

0

0,050

180

240

300

360

MCM v3a

Experimental

MCM v3

0,10

DTC387B

0,08

CTC014

0,045

0,09

0,040

0,08

0,035

0,07

0,030

0,06

0,05

0,025

0,05

0,04

0,020

0,04

0,015

0,03

0,010

0,02

0,005

0,01

0,000

0,00 60

120

Plots of experimental and calculated CH3CHO (ppm) vs. time (min) for the CH3CHO – NOx –air experiments.

Figure 5.

0

60

120

180

240

300

0,06

0,03 0,02 0,01 0,00

0

360

DTC150B

0,07

60

120

180

240

360

Experimental

MCM v3

SAPRC 99

300

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated HCHO (ppm) vs. time (min) for the CH3CHO – NOx –air experiments.

Figure 6.

0,012

0,05

DTC387B

0,010

CTC014

0,04

0,008 0,03

0,006 0,02

SAPRC 99 MCM v3 Experimental MCM v3a

0,004 0,01

0,002 0,000

0,00

0

60

Figure 7.

120

180

240

300

360

0

60

120

180

240

300

360

Plots of experimental and calculated PAN (ppm) vs. time (min) for the CH3CHO – NOx –air experiments.

MEK-NOX experiments The MCM v3 mechanism was found to over-predict D(O3-NO) in many of the chamber runs, particularly in the early stages of the experiment. This discrepancy was improved by reducing the quantum yield for the photolysis of MEK (reaction (7)). In MCM v3, a wavelength-independent value of 0.34 is applied, based on the atmospheric pressure data from Raber and Moortgat, (1996). That study demonstrated that the quantum yield decreases with increasing pressure, but also quoted increasingly wide uncertainty limits on the quantum

248

P. G. Pinho et al.

yield at pressures approaching one atmosphere. The value was therefore optimized within MCM v3a, to give the best fit to the MEK-NOx chamber data (see Figure 8 to Figure 12). This led to a best fit value of 0.17, which is close to the extremity of the uncertainty limit of the determination of Raber and Moortgat, (1996). This value is also consistent with the value of 0.15 obtained by Carter, 2000 during evaluation and optimization of the SAPRC-99 mechanism. 0,8

0,6

DTC337A

0,7

0,5

CTC178A

0,5

DTC361A

0,4

0,6

0,4

0,5

0,3

0,3

0,4

0,2

0,3

0,2

0,2

0,1

0,1

0,1

0,0

0,0 0

60

120

180

240

300

0,0

0

360

60

SAPRC 99

120

180

240

300

360

0

60

120

Experimental

MCM v3

180

240

300

360

MCM v3a

Plots of experimental and calculated D(O3-NO) (ppm) vs. time (min) for the MEK – NOx –air experiments.

Figure 8.

8,0

9,2

DTC337A

7,8

9,3

CTC178A

7,6

DTC361A

9,2

9,1

9,1

9,0

7,4

9,0

8,9

7,2

8,9

8,8

7,0

8,8

6,8

8,7

6,6

8,6

0

60

120

180

240

300

360

8,7 8,6

0

SAPRC 99

60

120

180

240

MCM v3

300

360

0

60

120

180

240

300

360

MCM v3a

Experimental

Plots of experimental and calculated MEK (ppm) vs. time (min) for the MEK – NOx –air experiments.

Figure 9.

0,10

0,12

DTC337A

0,08

CTC178A

0,09 0,08

DTC361A

0,07

0,10

0,06

0,07

0,08

0,05

0,06 0,05

0,06

0,04

0,04

0,03

0,04

0,03

0,02

0,02

0,02

0,01

0,01 0,00

0,00 0

60

120

180

240

SAPRC 99

Figure 10.

300

360

0,00

0

60

MCM v3

120

180

240

300

360

Experimental

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated HCHO (ppm) vs. time (min) for the MEK – NOx –air experiments.

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3 0,18

0,25

DTC337A

249

DTC361A

0,16

0,20

0,14 0,12

0,15

0,10 0,08

0,10

SAPRC 99 MCM v3 Experimental MCM v3a

0,06 0,04

0,05

0,02

0,00

0,00 0

60

120

180

240

300

0

360

60

120

180

240

300

360

Plots of experimental and calculated CH3CHO (ppm) vs. time (min) for the MEK – NOx –air experiments.

Figure 11.

0,16

0,14

DTC337A

0,14

0,09

CTC178A

DTC361A

0,08

0,12

0,07

0,12

0,10

0,06

0,10 0,08

0,05

0,06

0,04

0,08 0,06

0,03 0,04

0,04

0,02 0,02

0,02 0,00

0,01 0,00

0,00

0

60

120

180

240

300

360

0

SAPRC 99

Figure 12.

60

MCM v3

120

180

240

300

360

Experimental

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated PAN (ppm) vs. time (min) for the MEK – NOx –air experiments. The experimental PAN data in the CTC chamber runs appear anomalous and are not well described by either MCM v3a or SAPRC99.

Butane-NOX experiments Because there is no primary radical source during butane photo-oxidation, this parent system was the most sensitive to changes in the chamber radical source parameter. In conjunction with the optimizations described above, it was found that the best fit to the butane-NOX photo-oxidation data was obtained by multiplying the Carter chamber radical source values used in the SAPRC evaluation (Carter, 2000) by a factor of 1.35 for the ITC and ETC chambers, and by a factor of 1.2 for the DTC, CTC and XTC chambers. With these changes, the adapted MCM v3a butane mechanism was found to give generally good fits to D(O3-NO) data in the complete series of butane–NOX photo-oxidation experiments. The scatter in the results was indicative of run-to-run variability, with most of the data being fit by the mechanism to within +/- 30%, with no consistent biases. Examples of D(O3-NO), butane decay rate and formation of CH3CHO and MEK are presented in Figure 13 to Figure 16. As indicated above, the butane, HCHO, CH3CHO and MEK systems were considered iteratively, such that the optimized chamber radical sources and the changes made to the photolysis parameters for HCHO and MEK led to a self-consistent description of all the systems.

250

P. G. Pinho et al.

0,12

0,25

ITC939

0,10

0,30

ITC533

ITC507

0,25

0,20

0,20

0,08

0,15

0,15

0,06

0,10

0,10

0,04

0,05

0,02

0,05 0,00

0,00

0,00 0

60

120

180

240

300

0

360

SAPRC 99

60

120

180

240

300

0

360

60

120

Experimental

MCM v3

180

240

300

360

MCM v3a

Plots of experimental and calculated D(O3-NO) (ppm) vs. time (min) for the butane – NOx -air experiments.

Figure 13.

4,90

3,00

ITC939

3,80

ITC533

ITC507

4,85 2,95

3,75

2,90

3,70

2,85

3,65

4,80

4,75

4,70

4,65

2,80

0

60

120

180

240

300

360

3,60

0

SAPRC 99

60

120

180

240

MCM v3

300

360

0

60

120

180

240

300

360

MCM v3a

Experimental

Plots of experimental and calculated butane (ppm) vs. time (min) for the butane – NOx -air experiments.

Figure 14.

0,040

0,06

ITC939

0,035

0,07

ITC533

0,030

0,05

0,04

0,025

ITC507

0,06

0,05

0,04 0,020

0,03

0,03 0,015

0,02

0,02

0,010 0,01

0,005 0,000

0,01 0,00

0,00

0

60

120

180

240

SAPRC 99

Figure 15.

300

360

0

60

MCM v3

120

180

240

300

360

Experimental

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated CH3CHO (ppm) vs. time (min) for the butane – NOx -air experiments.

Evaluation of the Detailed Tropospheric Chemical Mechanism, MCM v3 0,030

0,045

ITC939

0,08

ITC533

0,040

ITC507

0,07

0,025

0,035

0,06

0,030

0,020

251

0,05

0,025 0,015

0,04

0,020 0,03

0,015

0,010

0,010

0,02

0,005

0,01

0,005

0,000

0,000 0

60

120

180

240

300

SAPRC 99

Figure 16 .

360

0,00

0

60

MCM v3

120

180

240

300

360

Experimental

0

60

120

180

240

300

360

MCM v3a

Plots of experimental and calculated MEK (ppm) vs. time (min) for the butane – NOx -air experiments.

Summary and Conclusions The MCM v3 butane degradation chemistry has been evaluated using chamber data on the photo-oxidation of butane, and on photo-oxidation of its degradation products, methylethyl ketone (MEK), acetaldehyde (CH3CHO) and formaldehyde (HCHO), in conjunction with an initial evaluation of the chamber-dependent auxiliary mechanisms for the series of relevant chambers. The MCM v3 mechanism for butane was found to provide an acceptable reaction framework for describing the NOx-photo-oxidation experiments on the above systems, although a number of parameter modifications and refinements were identified and introduced which resulted in an improved performance. These generally relate to the magnitude of sources of free radicals from carbonyl photolysis processes, which are currently under review in MCM development activities. Specifically, recommendations are made to update the photolysis parameters for HCHO, MEK, either in line with data reported since the MCM mechanism development protocol (Jenkin et al., 1997), or on the basis of optimization in the current study. Throughout the study, the performance of MCM v3 and the refined mechanism (denoted MCM v3a) was simultaneously compared with that of the SAPRC-99 mechanism, which was developed and optimized in conjunction with the chamber datasets (Carter, 2000). The performance of MCM v3a was found to be very similar to that of SAPRC-99. References Bloss C., V. Wagner, A. Bonzanini, M. E. Jenkin, K. Wirtz, M. Martin-Reviejo and M. J. Pilling; Evaluation of detailed aromatic mechanisms (MCM v3 and MCM v3.1) against environmental chamber data, Atmospheric Chemistry and Physics Discussions, 4 (2004a) 5683-5731. Bloss C., V. Wagner, M.E. Jenkin, R. Volkamer, W. J. Bloss, J. D. Lee, D. E. Heard, K. Wirtz, M. MartinReviejo, G. Rea, J. C. Wenger and M. J. Pilling; Development of a detailed chemical mechanism (MCMv3.1) for the atmospheric oxidation of aromatic hydrocarbons, Atmospheric Chemistry and Physics Discussions, 4 (2004b) 5733-5788. Carter, W. P. L. and F. W. Lurmann; Evaluation of a Detailed Gas-Phase Atmospheric Reaction Mechanism using Environmental Chamber Data, Atmospheric Environment 25A (1991) 2771-2806. Carter W. P. L., D. Luo, I. L. Malkina and D. Fitz,; The University of California, Riverside Environmental Chamber Data Base for Evaluating Oxidant Mechanisms: Indoor Chamber Experiments Through 1993, EPA/600/SR-96/078, U.S. Environmental Protection Agency, Research Triangle Park, NC (1995a).

252

P. G. Pinho et al.

Carter W. P. L., D. Luo, I. L. Malkina and J. A. Pierce; Environmental Chamber Studies of Atmospheric Reactivities of Volatile Organic Compounds. Effects of Varying Chamber and Light Source, Final report to National Renewable Energy Laboratory, Contract XZ-2-12075, Coordinating Research Council, Inc., Project M-9, California Air Resources Board, Contract A032-0692, and South Coast Air Quality Management District, Contract C91323 (1995b). Carter W. P. L.; Documentation of the SAPRC-99 Chemical Mechanism for VOC reactivity Assessment, Final Report to California Air Resources Board Contract 92-329 and Contract 95-308, Air Pollution Research Center and College of Engineering Center for Environmental Research and Technology University of California Riverside, California (2000). Derwent, R.G., M.E. Jenkin, S.M. Saunders and M.J. Pilling; Photochemical ozone creation potentials for organic compounds in North West Europe calculated with a master chemical mechanism, Atmospheric Environment, 32 (1998) 2429-2441. IUPAC Subcommittee on Gas Kinetic Data Evaluation, http://www.iupac-kinetic.ch.cam.ac.uk/. Jeffries H.E., M.W. Gery and W.P.L. Carter; Protocols for evaluating oxidant mechanisms for urban and regional models, EPA 600/R-92/112. Atmospheric Research and Exposure Assessment Laboratory, ORD, USEPA, Research Triangle Park, North Carolina (1992). Jenkin M.E., S.M. Saunders and M.J. Pilling; The tropospheric degradation of volatile organic compounds: A protocol for mechanism development, Atmospheric Environment, 31 (1997) 81-104. Jenkin M.E., T.J. Davies and J.R. Stedman; The origin and day-of-week dependence of photochemical episodes in the UK. Atmospheric Environment, 36 (2002) 999-1012. Jenkin, M. E., S. M. Saunders, V. Wagner and M. J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3, Part B: tropospheric degradation of aromatic volatile organic compounds. Atmospheric Chemistry and Physics, 3 (2003) 181-193. Killus, J. P. and G. Z. Whitten; Backgroud reactivity in smog chambers. Int. J. Chem. Kinet. 22 (1990) 547-575. Meller R. E. and G. K. Moortgat; Temperature dependence of the absorption cross sections of formaldehyde between 223 and 323 K in the wavelength range 225-375 nm, J. Geophys.Res. 105 (2000) 7089-7101. Raber W.H. and G.K. Moortgat; Progress and Problems in Atmospheric Chemistry, edited by J. Barker, 318-373, World Scientific Publ. Co., Singapore (1996). Saunders S.M., M.E. Jenkin, R.G. Derwent and M.J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3, Part A: tropospheric degradation of non-aromatic volatile organic compounds, Atmospheric Chemistry and Physics, 3 (2003) 161-180. Smith G. D., L. T. Molina and M. Molina; Measurement of Radical Quantum Yields from Formaldehyde Photolysis between 269 and 339 nm, J.Phys.Chem. A, 106 (2002) 1233-1240.

Studies on Nitrate-Affected SO2 Oxidation and Their Perspectives Wanda Pasiuk-Bronikowska and Tadeusz Bronikowski Institute of Physical Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw, Poland Key Words: Autoxidation mechanism, Cross-activation, Nitric radicals, Reactors, Sulphoxy radicals

Introduction Atmospheric chemistry has not yet been fully understood. One of the main reasons is the complex interaction between components of the gas phase and of the condensed phases. In this contribution we would like to consider, by example of the reaction between aqueous SO2 and molecular oxygen, occurring in the presence of nitrate, some aspects of such an interaction which may be of interest both to the researchers and modellers of cloud chemistry. Sources of nitrate in tropospheric droplets Let us first consider briefly the sources of nitrate in the troposphere aqueous phase. In modelling tropospheric multiphase chemistry, a number of more or less recognized gas phase and aqueous phase reactions are included which produce nitric species. We focus on the following main modes leading to nitrate within a droplet: Mode I. Dissolution of nitric species generated in a gas phase;

NO2 + OH + M ĺ HNO3 + M during daytime (Crutzen, 1995)

NO3 +

ĺ HNO3 +

The reaction between nitric radicals and pinonaldehyde (Waengberg et al., 1997) is only one example of the role of organic compounds in the gas phase formation of nitric acid in the atmosphere. At this point the heterogeneous reactions between gaseous species and solid or humidified NaCl should also be mentioned as a source of easily water soluble NaNO3: NO3 + NaCls ĺ NaNO3 + Cl ClONO2 + NaCls ĺ NaNO3 + Cl2 Mode II. Uptake of trace N-compounds, accompanied by a chemical reaction in a droplet: 2NO2 + H2O ĺ NO2Ǧ + NO3Ǧ + 2H+ N2O5 + H2O ĺ 2NO3Ǧ + 2H+ HNO4 + HSO3Ǧ ĺ NO3Ǧ + SO42Ǧ + 2H+ NO3 + org ĺ NO3Ǧ + org• 253 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 253–260. © 2006 Springer. Printed in the Netherlands.

254

W. Pasiuk-Bronikowska and T. Bronikowski

Reactions with water remove reactive nitrogen oxides from the gas phase and, moreover, transform nitrogen oxides into reservoir species, nitric acid and its anions. Reactions involving very active radicals generated in the gas phase, like NO3, are less probable in a droplet bulk. Because of the high activity of these radicals, their gas-phase concentrations and mass fluxes are relatively low, so that the radicals should rather become exhausted on the gas-liquid interface. The mechanism of such reactions given by Herrmann et al. (1996) for aromatic hydrocarbons clearly shows the various pathways leading to nitric acid. Mode III. Uptake of nitric acid during the condensation of water. The results obtained so far in field experiments (Laj et al., 1997) do not allow to establish whether nitrate molecules become incorporated within droplets by nucleation at the onset of cloud or fog or if they are scavenged by the droplets from the interstitial air. Nitrate anions are one of the major ionic species present in the Po Valley fog water drops (Decesari et al., 2001). Recently, the Black Sea was shown to release NO3Ǧ rich aerosols into the troposphere due to the bubble bursting process (Karakas et al., 2001). In general, aerosol nitrate is estimated to double in the next half century (Thiemens et al., 2003). Nitric acid is one of the so called “reservoir species” of the atmosphere (Barker, 1995). This means that one may expect the relatively easy release of temporarily “stored” nitric radicals when such a reservoir species is attacked by other radicals. Among these radicals are sulphate radical anions, SO4• Ǧ, formed as intermediates in the autoxidation of S(IV), in the oxidation of aqueous SO2 by molecular oxygen. We shall discuss consequences of this fact in the course of this contribution. It is why sulphur chemistry in the presence of nitrates becomes an area that needs to be further developed, as the cross-activation of these trace components may be important for understanding some field observations. Nitrate as an inhibitor of SO2 autoxidation to sulphuric acid The autoxidation of S(IV) is known to occur by a long chain mechanism (Table 1). Such a mechanism comprises three sequences: initiation, propagation and termination. It is essential to understand clearly that the initiation and termination steps have rates lower, by at least of an order of magnitude, than those of propagation steps. The intervention of nitrate anions in the mechanism of S(IV) autoxidation depends on the concentration of these anions. At relatively low nitrate concentrations, the rate of scavenging SO4• Ǧ radical anions attains low values, comparable with the rate of chain termination. In this instance two specific cases may be distinguished (Table 2): Ɣ The radical product of the reaction between nitrate and SO4• Ǧ (NO3) attains too low concentrations to influence the overall reaction kinetics. This is the classic case of inhibition due to a reaction in which reactive radicals are replaced by other ones, which are inactive under the conditions of the experiments. Ɣ The accumulation of NO3 radicals becomes essential, so that the activity of these radicals is noticeable. The reacting system is switched to the case when the secondary initiation emerges, thus reducing the effect of the nitrate inhibitor. The net effect is the retardation of the rate of S(IV) autoxidation.

Studies on Nitrate-Affected SO2 Oxidation and Their Perspectives

255

Classic mechanism of a long chain reaction.

Table 1. initiation:

-

I

-

MIII + SO32 / HSO 3

-

-

MII + SO3 /( SO3 + H+)

propagation: SO3 + O2 SO5 kp1 SO5 + SO32 /HSO 3 SO52 /HSO5 + 2 SO5 + SO3 /HSO 3 SO42 /HSO4 + 2 SO4 + SO3 /HSO 3 SO42 /HSO4 + nonradical S(IV) oxidation: SO52 + SO32 /HSO 3 2SO 42 /(2SO 42

- - - - termination: 2SO S O -+ O

-

I

I I I

I

I

5

2

2 8

2

-

ki

-

SO 3 (rate controlling step) SO4 kp3 SO 3 kp4

-+ H

+

)

kp2

kp5

(nonradical products )

kt

I

Interference of nitrates (low concentration).

Table 2.

Termination (classic): 2SO5 S2O82 + O2

I

-

-

kt

Termination (nitrate scavenger): NO3 + SO4 NO3 + SO42 ksc (radical product not effective kinetically)

I

-

-

-

Retardation (nitrate scavenger): NO3 + SO4 NO3 + SO42 (radical product kinetically effective)

I

-

-

Initiation (primary ): MIII + HSO3 MII + SO3 + H+ (secondary): NO3 + HSO3 NO3 + SO3 + H+

I I

-

-

-

-

-

k sc

ki,p ki,s

Two plots are shown in Figure 1, the upper one indicates the concentration of S(VI) (the S(IV) autoxidation product) growing with the time of reaction performed under quasisteady state conditions and the lower one which gives the rate - scavenger (nitrate) concentration profile. It is evident that the rate of S(IV) autoxidation is a non-monotonous function of the concentration of nitrate. The question arises, what is the influence of nitrate on the rate of S(IV) autoxidation at still higher concentrations of this compound.

W. Pasiuk-Bronikowska and T. Bronikowski

256

' [ SVI ]

In this light the question arises what will happen at still higher concentrations of nitrate; should one expect the total compensation of the inhibition by nitrate as observed at low concentration, when the nitrate concentration attains a sufficiently high value? 1

3

Figure 1. Effect of nitrates (low concentration region) on the rate of S(IV) autoxidation: accumulation of the reaction product, S(VI), in time (upper plots); rate – scavenger (nitrate) concentration profile.

2

time

r

?

z1 3

z2

z

Cases of the intervention of nitrate: 1. No nitrate in the reacting system; 2. Nitrate present at very low concentration (no effect of nitric radicals on chain initiation); 3. Nitrate present at low concentration (substantial effect of nitric radicals on chain initiation).

> Sc@ Nitrate as a catalyst of SO2 autoxidation to sulphuric acid Sea aerosols initially have an ionic composition of sea water but quickly loose water to attain equilibrium with water vapour in the surrounding air, simultaneously cumulating reactive trace gases and undergoing chemical reactions (Pszenny et al., 1998). Also cloud droplets may undergo dehydration, depending on air humidity. The concentration of nitrate may reach very high values, a mole per dm3 and over, when droplets evaporate almost to dryness. High concentrations of nitrate should also be expected in the case when initially dry or humidified solid NaCl aerosols are reacted with nitric radicals, NO3, or chlorine nitrate, ClONO2, reactions recently discussed, for instance, by Gershenzon et al. (2002). Our recent study on the effect of nitrate at high concentrations led to the results which give ground for inferring that the intervention of nitrates may be also at the level of chain propagation (Pasiuk-Bronikowska et al., 2004). Then, the rate of S(IV) autoxidation substantially exceeds that attained in the absence of nitrate. This is illustrated in Figure 2. When the concentration of nitrate attains a sufficiently high level, the reaction rate, r, starts to grow above the value for the uninhibited reaction. Such an acceleration of S(VI) formation in the presence of high amounts of nitrate indicates that the latter compound must react in the rate determining step. As in this part of our experiments the excess of nitrate with respect to S(IV) was immense, it seemed reasonable to explain the observed changes in the S(IV) autoxidation kinetics assuming that:

Studies on Nitrate-Affected SO2 Oxidation and Their Perspectives

257

Ɣ the presence of sulphate radicals in the reacting system is negligible, Ɣ the hydrogen sulphite anions in the rate controlling step, suitable in the case of low concentrations of nitrate, are replaced by nitrate anions.

Effect of nitrates (concentration region extended) on the rate of S(IV) autoxidation.

Figure 2.

As shown in Table 3, nitrate anions consumed in the rate controlling step are reformed in the next step involving nitric radicals and S(IV). This is a sort of a catalytic cycle, where Interference of nitrates (high concentration).

Table 3. initiation:

-

propagation: SO3 + O2

I

MIII + HSO3–

-

MII + •SO3 – + H+

k’i

SO k (rate controlling step) SO - + NO SO - + NO NO + HSO - NO - + SO - + H k nonradical S(IV) oxidation: SO - + HSO 2SO - + H k I I I

•–

5

p1

–

5

3

– 3

2– 5

– 3

I

3

5

– 3

•

2– 4

3

+

3

+

p

p5

termination: 2SO5 S 2O82 + O2 (nonradical products )

I

-

-

I

2–

kt

W. Pasiuk-Bronikowska and T. Bronikowski

258

nitrate anions play a role of a catalyst. The location of point 4 (white spot in Figure 2) is evidence for the nitrate catalysis. These surprising results shed light on the possible important role of nitrate in the atmospheric autoxidation of SO2 by molecular oxygen. In a certain sense they coincide with the conclusions derived by Leriche et al. (2000) who reported on a multiphase box model for a remote region of the troposphere applied to a set of field measurements in France. It is commonly believed that the main oxidant at low pH is H2O2. To account for the observed conversion of S(IV) to S(VI) at very low concentrations of H2O2 Leriche et al. postulated the decisive role of peroxy nitric acid, HNO4. These authors included in their model of Nchemistry in droplets the following reaction: NO2 + HO2/O2Ǧ ĺ HNO4/ NO4Ǧ According to Amels et al. (1996), the S(IV) oxidation by peroxy nitric acid can proceed very rapidly with the simultaneous formation of nitrate: HNO4 + HSO3Ǧ ĺ NO3Ǧ + SO42Ǧ + 2H+ Both oxidation pathways, the NO3Ǧ catalysed oxidation of S(IV) by molecular oxygen and the oxidation of S(IV) by peroxy nitric acid lead to the formation of sulphate anions and further studies are required to decide, which of the two pathways is the mot important (or perhaps both) under real atmospheric conditions. A chamber reactor as a tool for scaling up A variety of reactor types are known which may be roughly divided into two categories: homogeneous and heterogeneous ones, depending on whether the studied reaction involves reactants present in a single phase or whether they must be drawn from two or more phases. The autoxidation of SO2 is an instructive example of a reaction which leads to different kinetic observations depending on the way it investigated in a particular reactor. In the case of a single-phase laboratory reactor, shown schematically in Table 4, the reactor volume is totally occupied by liquid. The autoxidation of SO2 occurs between reactants, including oxygen, pre-dissolved in the liquid. Thus, the only reaction rate limitations come from the properties of a measuring device (Clark-type electrode) and from the exhaustion of reactants. In a two-phase reactor, with liquid as a continuous phase, the reaction may be observed for a longer time, as the gas reactant is continuously replenished. Here, the intrinsic reaction rate limitation becomes the rate of oxygen supply across the oxygen/water interface (including that of gas bubbles). In a three-phase reactor, also with liquid as a continuous phase, the oxygen/water (slurry) interface is planar, what allows the determination of the rate of oxygen diffusion. A solid suspension is a source of the reacting solute supplied by dissolution. Usually, in a chamber reactor (see the bottom row in Table 4) the complex chemical systems existing in the atmosphere gas phase are approximated. Especially advantageous for studying gas reactions are the two characteristic features of a chamber reactor: a relatively large volume and rather insignificant wall effects. Examples of effectively studied tropospheric reactions in chamber reactors are: photo-dissociation and oxidation of a selection of organic compounds, the latter reactions with such oxidants as ozone, OH and NO3. In many

Studies on Nitrate-Affected SO2 Oxidation and Their Perspectives

259

instances, the studies were focused on gas-phase reaction mechanisms and, to a lesser extent, on the gas-phase reaction kinetics. With the introduction of water droplets or wet aerosol particles into the interior of a chamber, mass transfer in the sense of microphysics emerges as a coordinate part of the problem of the proper approximation of chemical transformations in such a system. To quantitatively describe the mass transfer/chemical reaction interplay is not a trivial task. It requires a set of data on gas-side and liquid-side diffusion to be collected and included in existing models.

Applicability of some typical reactors.

Table 4.

TYPE

REACTOR SCHEME oxygen sensor

liquid-phase (single-phase reactor)

oxygen inlet

gas-liquid (two-phase reactor)

gas-liquid-solid (multiphase reactor)

pH sens or or oxygen probe

air inle t outle t conductometric probe

solid suspension

gas-phase

instrume ntation instrumentation

gas-liquid gas-liquid-solid

instrumentation

REACTOR CHARACTERISTICS • volume: 4x10-3 dm3; • reaction time: several sec; • gas reactant predissolved; • formation of secondary products avoided; • serves to determine kinetic laws and rate constants; • volume: 0.10 dm3; • reaction time: tens of min; • gas reactant supplied by diffusion; • formation of secondary products avoided; • serves to determine kinetic laws and rate constants;

• volume: 0.15 dm3; • reaction time: tens of min; • gas reactant supplied by diffusion; • solute reactant supplied by solid dissolution; • formation of secondary products avoided; • serves to determine kinetic laws and rate constants;

• volume: up to ca. 2x10 5 dm3 (200 m3); • reaction time: hours to days; • formation of secondary products allowed; • serves the determination of chemical composition changes in time (product identification and distribution); • influence of mass transfer (convection/diffusion)

?

We would like to raise the question of the inconveniency resulting from the complexity of cloud chemistry models, at least when taking into account the required computing power and time. To simplify the model approach we suggest treating a cloud not as an aqueous phase volume, but as a system of mixed phases, working as a whole. It would require confining a cloud in a chamber reactor and determining new kinetic parameters as if it were a somewhat modified gas-phase. Conclusions The widely held view that reactions of relatively low rate constants may be safely neglected in the modelling of aqueous phase chemistry should be verified. As shown for the cross-activation of S(IV) and nitrate, the neglected reactions may accelerate, possibly by orders of magnitude, due to an increase of the concentration of inorganic components in desiccating aerosols.

260

W. Pasiuk-Bronikowska and T. Bronikowski

The use of a chamber reactor for “confining a cloud in a cage” and treating it as a modified gas phase is suggested, here, as an approach to reasonably simplified modelling of chemical composition changes in real clouds. A comparison between the results obtained using a large volume chamber reactor and those calculated from the aqueous phase model, based on data collected using small laboratory reactors, should shed light on the importance of gas-side and liquid-side resistances to mass transfer. Acknowledgements This work, being the contribution to the ESF programme INTROP, is a logical continuation of our study developed within the frames of he EUREKA project, EUROTRAC2. The financial support from the Scientific Committee for Research in Poland is gratefully acknowledged. References Amels, P., H. Elias, U. Goetz, U. Steingens and K. J. Wannowius; Kinetic investigation of the stability of peroxonitric acid and of its reaction with sulphur(IV) in aqueous solution, in: P. Warneck (ed.), Heterogeneous and Liquid-Phase Processes, Springer, Berlin (1996) 77-88. Barker, J. R.; A brief introduction to atmospheric chemistry, in: J. R. Barker (ed.), Progress and Problems in Atmospheric Chemistry, World Scientific, London (1995) 1-33. Crutzen, P. J.; Overview of tropospheric chemistry: developments during the past quater century and a look ahead, Faraday Discuss., 100 (1995) 1-21. Decesari, S., M. C. Facchini, E. Matta, M. Mircea, S. Fuzzi and J.-P. Putaud; Organic and inorganic solutes in atmospheric aerosol: a full characterisation approach, in: P.M.Midgley, M.J.Reuter, M.Williams (eds), Proc. EUROTRAC-2 Symposium 2000 “Transport and Chemical Transformation in the Troposphere”, Springer, Berlin (2001) 274-277. Gershenzon, M. Yu., V. M. Grigorieva, S. D. Il’in, R. G. Remorov, D. V. Shestakov, V. V. Zelenov, E. A .Aparina and M. Yu. Gershenzon; Heterogeneous reactions affecting chlorine activation in the troposphere, in: I. Barnes (ed.), Global Atmospheric Change and its Impact on Regional Air Quality, Kluwer Academic Publishers, Dordrecht (2002) 109-113. Herrmann, H., H.-W. Jacobi, G. Raabe, A. Reese, Th. Umschlag and R. Zellner; Free radical reactions in the tropospheric aqueous phase, in: B. Larsen, B. Versino, G. Angeletti, (eds), The Oxidizing Capacity of the Troposphere, Report EUR 17482 (1997) 503-512. Laj, P., S. Fuzzi, L. Ricci, A. Berner, D. Schell, M. Wendisch, W. Wobrock, G. Frank and B. Martinssonn; Chemical inhomogeneities in fog droplets during the 1994 CHEMDROP experiment, in: B. Larsen, B. Versino, G. Angeletti (eds), The Oxidizing Capacity of the Troposphere, Report EUR 17482 (1997) 513521. Karakas, D. and S. G. Tuncel; Long-range transport of aerosols to Turkish western Black Sea region, in: P.M.Midgley, M.J.Reuter, M.Williams (eds), Proc. EUROTRAC-2 Symposium 2000 “Transport and Chemical Transformation in the Troposphere”, Springer, Berlin (2001) 290-293. Leriche, M., D. Voisin, N. Chaumerliac, A. Monod and B. Aumont; A model for tropospheric multiphase chemistry: application to one cloudy event during the CIME experiment, Atmos. Environ., 34 (2000) 5015-5036. Pasiuk-Bronikowska, W., T. Bronikowski and M. Ulejczyk; Intervention of inorganic radical scavengers during the autoxidation of S(IV), Environ. Sci. Technol., submitted for publication. Pszenny, A., W. Keene, C. O’Dowd and M. Smith; Sea salt aerosols, tropospheric sulfur cycling, and climate forcing, IGACtivities Newsletter (11) (1998) 6-12. Thiemens, M., G. M. Michalski, A. Romero and J. R. McCabe; Mass independent oxygen and sulfur isotopic compositions of environmental sulfate and nitrate. A new probe of atmospheric, hydrospheric and geological processes, Geophys. Res. Abstracts 5 (2003) no. 04359. Waengberg, I., I. Barnes and K.-H. Becker; NO3 radical initiated oxidation of Į-pinene, in: B. Larsen, B. Versino, G. Angeletti (eds), The Oxidizing Capacity of the Troposphere, Report EUR 17482 (1997) 104-108.

Heterogeneous and Aqueous-Phase Transformations of Isoprene Krzysztof J. Rudzinski Institute of Physical Chemistry of the PAS, Kasprzaka 44/52, 01-224 Warsaw, Poland Key Words: Isoprene, Troposphere, Feedback, Sulphate Radical, Oxidation

Introduction Isoprene is a conjugated diene (2-methyl-buta-1,3-diene), volatile and hardly soluble in water; under normal pressure it boils at 34 oC (Merck, 1999) and dissolves up to 1.47×10-2 M at 21.5 oC, with a Henry’s constant of 0.027 mole kg-1 atm-1 at 25 oC (NIST, 2001). Isoprene is a metabolite in plants, microbes, animals and humans, and a major biogenic trace compound emitted to the atmosphere. It is very reactive towards atmospheric gas-phase oxidants such as hydroxyl and nitrate radicals or ozone. At higher concentrations, 220 – 7000 ppm, it is carcinogenic to rodents and possibly carcinogenic to humans (Melnick and Sills, 2001). Terrestrial vegetation produces most of the emitted isoprene – 500 Tg C per year (Guenther et al., 1995). The emitting plants synthesise isoprene in plastids (chloroplasts) from pyruvate and glyceraldehyde-3-phosphate, most probably to better resist the thermal stress (Harley et al., 1999; Kesselmeier and Staudt, 1999; Logan et al., 2000; RodriquezConcepcion and Boronat, 2002). However, many plants do not synthesise isoprene at all. Other natural sources of isoprene include sea phytoplankton (1 Tg C per year, Matsunaga et al., 2002), microbes (Wagner et al., 1999; Fall and Copley, 2000), animals and humans (Fenske and Paulson, 1999; Diskin et al., 2003). In humans, the synthesis of isoprene takes a mevalonate path that is related to the synthesis of cholesterol (Karl et al., 2001). Isoprene exhaled by humans totals roughly 4 Tg C per year. Anthropogenic sources of isoprene, which generally are poorly quantified, include combustion and evaporation of fuels (Broderick and Marnane, 2002; Borbon et al. 2003a, 2003b), cigarette smoking (Pouli et al., 2003; Baek and Jenkins, 2004), biomass burning (Andreae and Merlet, 2001), garden waste (Wilkins and Larsen, 1996) and chemical technology (Leber, 2001). Potential health risks induced the research interest in emissions of isoprene with vehicle exhaust and with cigarette smoke, as well as from facilities manufacturing natural and synthetic rubber, where concentrations of isoprene in ambient air are relatively high. Isoprene is not stored by organisms or in the environment; it encounters sinks wherever it is synthesised or transported, if only reactants are available. Endogenous isoprene may react with oxidants in the intracellular and extracellular fluids, or at walls and membranes in living organisms (Wannaz et al., 2003). The plausible reactants are oxidative species, such as free radicals or radical ions, or ozone. Expelled through the leaf stomata, or with human or animal breath, isoprene may react at the interface or in the volume of fluid films covering the body surfaces; leaves, membranes or skin. Unreacted isoprene is transported to the atmosphere and consumed in several gas-phase reactions (Atkinson and Arey, 2003, 2003a). These reactions are considered a main sink for isoprene in the environment. The possibility of heterogeneous reactions of isoprene on airborne aerosols and of multiphase reactions in atmospheric waters is discussed in this overview. Isoprene may also react in seawater, where 261 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 261–277. © 2006 Springer. Printed in the Netherlands.

K. J. Rudzinski

262

it is produced, or at the sea/air interface. Up to now, only the concentration of isoprene in seawater and the fluxes of isoprene desorbing from the sea to the atmosphere have been measured, while the possibility of aqueous-phase reactions has only been suggested (Milne et al., 1995; Matsunaga et al., 2002). Little is known on the chemistry of isoprene in soils, which naturally provide opportunities for gas-phase, heterogeneous and multiphase reactions. Some soil microbes are a well-quantified sink for isoprene with recognised biochemistry (Boyd et al., 2000), while other microbes produce isoprene (Wagner et al., 1999). The overall balance of isoprene exchange with soils has been estimated at 20.4 Tg C y-1, in favour of deposition (Cleveland and Yavitt, 1997). Gas-phase transformations of isoprene The gas-phase chemistry of isoprene is beyond the scope of this overview, but a short summary is given for completeness, and because many of the gas-phase descendants of isoprene can condense, dissolve in water or react on surfaces under atmospheric conditions.

OH

•

+ O2

OH

+

NO3

•

ONO2 O3

Figure 1.

OO • OH

+ O2

OO •

ONO2

O O O

Gas-phase reactions of isoprene with atmospheric radicals and ozone (for simplicity, only one isomer of each product is shown).

Isoprene does not photolyse in the atmosphere, but reacts with hydroxyl or nitrate radicals and, successively, with molecular oxygen, or with ozone (Figure 1). A hydroxyl radical adds to either of two double bonds in the isoprene molecule to give a hydroxyalkyl radical (four isomers are possible). Hydroxyalkyl radicals react with molecular oxygen to give hydroxyalkyl peroxy radicals in seven isomeric forms. Nitrate radicals react in a similar manner to yield nitrooxyalkyl and nitrooxyalkyl peroxy radicals, each in eight isomeric forms. Ozone adds to either of double bonds of isoprene yielding two isomeric ozonides. Peroxy radicals can react with other radicals, such as NO, HO2 or RO2. A broad spectrum of possible products comprises hydroxy hydroperoxides, diols, hydroxycarbonyls, methyl vinyl ketone, methacrolein, 3-methylfuran, various nitrates (nitroxy peroxy-, hydroxy- and carbonyl-), unsaturated dinitrates, methyl glyoxal, glyoxal, glycol aldehyde, peroxyacetyl nitrate (PAN) and methacrolyl nitrate (MPAN). Ozonides decompose into Criegee radicals and formaldehyde, methyl vinyl ketone or methacrolein. The ultimate products of gas-phase transformation of isoprene in the atmosphere are carbon monoxide and carbon dioxide. Extensive summaries of the gas-phase chemistry of isoprene can be found in the Master Chemical Mechanism (Saunders et al., 2003; http://www.chem.leeds.ac.uk/Atmospheric/

Heterogeneous and Aqueous-Phase Transformations of Isoprene

263

MCM/ mcmproj.html), or in several reviews (Le Bras et al., 1997; Atkinson, 2000; Atkinson and Arey, 2003, 2003a; Fan and Zhang, 2004). Heterogeneous transformations of isoprene The research interest in heterogeneous transformations of isoprene developed recently, in response to the discovery of humic-like substances (HULIS) in atmospheric aerosols from rural and urban regions (Havers et al., 1998). HULIS are macromolecular or polymeric substances that have dark yellow to brown colour and strongly absorb UV and VIS light (Limbeck et al., 2003; Feng and Möller, 2004). They dissolve in water and contain polysaccharidic and aliphatic substructures (Havers et al., 1998), as well as mono-, di- and polycarboxylic acids (Decesari et al., 2002; Chan and Chan, 2003; Gelenceser et al., 2003; Kiss et al., 2003). Originally, no gas-phase mechanisms could explain the formation of HULIS in the atmosphere, so researchers turned to the heterogeneous transformations of isoprene as a possible explanation. Furthermore, the modellers suggested looking for heterogeneous sinks for isoprene, because the global balance of isoprene did not close in their models (von Kuhlmann et al., 2004). Limbeck et al. (2003) tested the “heterogeneous” hypothesis in a simple laboratory experiment. They passed a stream of synthetic air containing 200 – 2000 ppbv of isoprene through a quartz-fibre bed impregnated with sulphuric acid (50 – 3600 ng). The gas flow rate was 400 dm3 h–1 and the average residence time of isoprene in the bed was 10 ms. The authors varied the acidity of the quartz bed by replacing part of H2SO4 with (NH4)2SO4. In further experiments, they added ozone to the gas stream as a competitive gas-phase reactant. The yellowish products formed on the quartz bed were in part analysed by a thermogravimetric method, and in part extracted with water and, subsequently, with methanol for further analysis. The thermograms indicated the presence of low volatile products, while the water-extracts strongly absorbed the UV light. Methanol extracts were purified by solid phase extraction, from inorganic and polar organic compounds and dried to orange coloured residues. Diffusion reflectance infrared Fourier transform (DRIFT) spectra of these residues revealed the carboxylic, hydroxyl and carbonyl functionalities present. The formation rate of the products decreased with decreasing acidity of the bed, or when ozone was added to the inlet gas stream. Based on these results, Limbeck et al. (2003) suggested that the polymeric substances were formed in a heterogeneous reaction of isoprene with sulphuric acid. Moreover, they showed that the DRIFT spectra of the water-insoluble fractions of their products were extremely similar to spectra of the reference NIST dust. Czoschke et al. (2003) used two large Teflon bags (500 dm3 each) to study the formation of secondary aerosols from isoprene in the presence of ozone, at 22 – 27 oC. They filled the bags with clean, dry air containing 0.57 – 0.59 ppm of O3, and added 5 – 50 µg m–3 of seed aerosol prepared by mixing in various proportions and drying the 6.7 mM solution of (NH4)2SO4 and 5 mM solution of H2SO4. Then, they introduced gaseous isoprene or isoprenealcohol (nonanol or decanol) mixtures into the bags and measured the concentration and size distribution of particles forming, as well as the post-reaction concentrations of ozone. The authors calculated the aerosol yield as the ratio of the volume concentration of aerosol produced to the volume of reactive organic gas consumed. The latter value was estimated from the amount of ozone available to the system and the amount of OH radicals that would be produced in the isoprene-ozone reaction. Table 1 shows the initial concentrations of reactants and the calculated yields of aerosol. The amount of seed aerosol never exceeded

K. J. Rudzinski

264

0.02 to 0.5 % of the total aerosol formed. The yields were higher for acidic seed aerosol and in the presence of alcohols. Table 1. Aerosol yields in seeded ozone-isoprene-alcohol systems (Czoschke et al., 2003). Initial concentrations, ppm: Ozone Isoprene Alcohol Temperature, oC Seed aerosol Aerosol yield, %

3.7 no 24 neutral acidic 0.2 0.6

0.57-0.59 3.1 3.6 (nonanol) 27 neutral acidic 0.4 1.1

4.4 1.4 (decanol) 22 neutral acidic 0.7 1.6

The FTIR spectra showed that the aerosols formed contained hydrates, polymeric forms, acetals and hemiacetals. Several peaks in these spectra closely corresponded to peaks in FTIR spectra of aerosols collected in the Smoky Mountains (Jang et al. 2002). Czoschke et al. (2003) concluded, that in their experiments, isoprene had been oxidised to aldehydes by ozone in the gas phase, and the aldehydes had been converted to the observed products in heterogeneous, acid-catalysed reactions. However, the presented analysis of aerosol yields was very approximate, and did not exclude the possibility of heterogeneous reactions of isoprene in the discussed experiments. All the differences in reactors and techniques given, in the experiments by Limbeck et al. (2003) ozone reacted with isoprene in the gas phase, but decreased the amount of polymers formed. Thus, the heterogeneous reactions of isoprene can be a source of several components of atmospheric aerosols. The evidence is limited, however, and has to be supported by further experimental work. Other work has shown that the formation of polymeric compounds is also possible in non-acid or unseeded gas-phase experiments. Aqueous-phase transformations of isoprene Interest in the aqueous-phase reactions of isoprene first stemmed from attempts to clean the effluents from the natural rubber industry. Elkanzi and Bee Kheng (2000) studied the oxidation of isoprene in a small, stirred photoreactor (0.25 dm–3 volume, maximal lightsource intensity of 2.25 W m–2 at 254 nm and 75 mm distance). They irradiated aqueous solutions containing 1.47 mM of isoprene and 0 – 9 mM of H2O2 (pH = 6.8), for 1 to 3 hours, at 24 oC. The concentrations of isoprene and H2O2 in the reacting solutions were followed offline, using UV spectrophotometry (254 nm) and classical wet-chemical analysis methods. The rate of isoprene degradation was evaluated and best described by Equation 1: 

d >isoprene@ dt

k1>isoprene@  k 2 >isoprene@ >H2O2 @

(1)

The rate constants k1 = 3.67×10 – 4 s – 1 and k2 = 4.77×10 – 2 M – 1 s – 1 depended, in an unresolved way, on the intensity of irradiating light. To determine the reaction products, the authors extracted a few samples of the reacting solutions with ethyl acetate, acidified and dried the extracts, and analysed them with GC/MS. At short reaction times, they observed peaks corresponding to isoprene, formaldehyde and methyl vinyl ketone. In experiments with the highest initial ratio of H2O2 to isoprene, these peaks disappeared after 2 hours.

Heterogeneous and Aqueous-Phase Transformations of Isoprene

265

Pedersen and Sehested (2001) investigated the kinetics of the reaction of isoprene with ozone, being inspired by the smell of wet linen drying in the sun that was commonly attributed to ozone, but might originate from dilute products of isoprene oxidation – methacrolein and methyl vinyl ketone. These authors used a stopped-flow apparatus to mix the 47.1 µM aqueous solution of isoprene with a 6 µM solution of O3, at 5 – 30 oC and pH = 2 (the solution pHs were adjusted with HClO4, for stabilisation of ozone). They followed the extent of reaction in time by measuring the absorption of UV light by ozone, at 260 nm. Similar experiments were carried out for reactions of methacrolein or methyl vinyl ketone with ozone (initial concentrations were 59.5 µM and 57.8 µM, respectively). The reactions were first order with respect to ozone and first order with respect to the organic reactant. The rate constants and energies of activation determined by Pedersen and Sehested (2001) are shown in Table 2. Based on the low-resolution UV spectra of the reacting isoprene-ozone mixtures, these authors suggested a rapid formation of an intermediate, which could be a stabilised carbonyl oxide (Criegee radical) known from the mechanism of the gas-phase ozonolysis of isoprene. Table 2.

Second-order rate constants for reactions of isoprene, methyl vinyl ketone (MVK) and methacrolein (MACR) with ozone, determined by Pedersen and Sehested (2001).

Isoprene + ozone MACR + ozone MVK + ozone

Energy of activation

Rate constant

Reaction

M

–1

s

–1

kJ mol – 1

5 oC

30 oC

25 oC

5 – 30 oC

2.23×105 1.17×104 2.6×104

4.64×105 2.94×104 5×104

(4.2±0.2)×105 (2.4±0.1)×104 (4.4±0.2)×104

19.9±0.5 23.9±0.5 18.0±0.5

Very recently, Claeys et al. (2004, 2004a) discovered two isomers of 2-methyl-butane1,2,3,4-tetraols in aerosol samples collected over the Amazon rain forest and over the continental K-Puszta forest in Hungary. The samples also contained carboxylic acids and polysaccharidic compounds, such as malic acid, oleic acid, palmitic acid, stearic acid, arabitol, mannitol, glucose and levoglucosan. Originally, the authors suggested that tetraols were formed in a series of gas-phase reactions starting from the oxidation of isoprene by OH and O2 radicals, and just condensed to aerosols (Claeys et al., 2004). Such a mechanism required low NOx conditions, but in subsequent field campaigns, more tetraols were found at high NOx conditions (Claeys et al., 2004a). Thus, the authors hypothesised that tetraols were formed in acid-catalysed reaction of isoprene with hydrogen peroxide, presumably heterogeneously or in the aqueous-phase. They carried out simple, experiments at room temperature, in which they mixed vigorously 50 µl of isoprene, 50 µl of 50% aqueous solution of H2O2 and 25 µl of 0.1 M aqueous solution of H2SO4. The mixtures were acidic, with pH = 2. After two hours, the solutions were analysed by GC/MS, using trimethylsilyl derivatisation.. The only products found were the erythro and threo isomers of 2-methyl-butane-1,2,3,4-tetraol. The tetraols peaks were identified in GC/MS chromatograms obtained from aerosols collected in the field campaigns and processed in the same way as the laboratory samples. In addition, Claeys et al. (2004a) carried out similar experiments in which they replaced isoprene with methacrolein or methacrylic acid, the products of the gas-phase oxidation of isoprene. In both cases 2,3-

K. J. Rudzinski

266

dihydroxymethacrylic acid was produced, as well as some precipitate that has not been analysed yet.

1.0

initial concentration of isoprene, mM:

[O2] [O2]o

0.0319 0.0200 0.0040 0

0.5

(a) 0.0

0

1

2

3

4

5

-3

time × 10 , s

1.0

0.0814 0.0407 0.0235 0.0118 0.0061 0.0024 0

initial concentration of isoprene, mM:

[O2] [O2]o 0.5

(b) 0.0 0

2

4

6

8

-3

time × 10 , s

Figure 2.

Decay of oxygen during autoxidation of Na2SO3 initiated by thermal decomposition of K2S2O8 (a) or by Mn(III) (b), and inhibited by isoprene. Initial concentrations of reactants were: K2S2O8 1 mM or MnSO4 0.01 mM; Na2SO3 1.02 mM; O2 0.22-0.25 mM; for isoprene –

Our own interest in the aqueous-phase chemistry of isoprene stemmed from the research on the inhibition of autoxidation of SO2 and its possible impact on the long-range transport of this pollutant (Pasiuk-Bronikowska et al., 2000). Addition of dissolved isoprene slowed down the homogeneous autoxidation of sodium sulphite solutions initiated by thermal decomposition of K2S2O8 (Rudzinski and Pasiuk-Bronikowska, 2001) or catalysed by MnSO4 (Rudzinski et al., 2000). The experiments were performed in a stirred tank reactor operated in a batch manner, at 25 oC. The reactor contained no gas phase and was sealed to prevent any gas-liquid mass exchange. The reacting solutions were slightly alkaline, with pH = 8.0 – 8.5.

Heterogeneous and Aqueous-Phase Transformations of Isoprene

267

We followed the extent of autoxidation by measuring the concentration of dissolved oxygen versus the reaction time (Figure 2). The plots and the formulas approximating the rates of autoxidation (Eqs 2, 3) clearly show the inhibiting role played by isoprene.



d >O 2 @ k1 >O 2 @ k a1 >O 2 @ dt 1  k >isoprene @0

(for K2S2O8 experiments)

(2)



d >O 2 @ ka 0 dt

(for MnSO4 experiments)

(3)

k0 1  k c >isoprene @0

Further experiments were carried out to explain the kinetics of isoprene degradation during the autoxidation of S(IV) (Rudzinski, 2004). The reactor was equipped with a closed sampling loop for recording the UV spectra of reacting solutions in time. The spectra were decomposed to the component spectra of isoprene, Na2SO3 and K2S2O8, and the time traces of the reactant concentrations were determined (Figure 3). The results were analysed using the chain mechanism of S(IV) autoxidation to which a reaction of isoprene with SO4• – radicals was added:

CH 2 C(CH 3 )CHCH 2  SO x4 o products,

k 2,iso

(I)

The lower limit of the rate constant k2,iso for reaction (I) was estimated under a few quantitative approximations, and the upper limit required the pseudo-stationary-state approximation for SO4• – radicals and SO52 – ions (Egn 4). (7 .07 r 1 .22 ) u 10 8 M1s 1 d k 2, iso d (2 .12 r 0 .37 ) u 10 9 M1s 1

(4)

The real value of k2,iso was anticipated close to the upper limit. The errors given in (4) are just standard errors and do not reflect the errors that could originate from the assumption that isoprene reacted exclusively with sulphate radicals. Approximate analysis showed that if isoprene reacted also with sulphite radicals, the upper limit in (4) is overestimated by a few percent, and if isoprene reacted also with peroxymonosulphate radicals as well, the upper limit is overestimated by not more than 30%. The tentative mechanism of isoprene transformation accompanying the autoxidation of S(IV) was suggested, which started with the addition of a sulphate radical to a double bond in isoprene, followed by addition of oxygen and formation of a peroxy radical. The peroxy radical reacted with sulphite ion to the alkoxy radical, which in turn reacted either with oxygen or with a sulphite ion to give respectively trans-3-methyl-4-sulphoxy-2-buten-1-al or trans-3-methyl-4-sulphoxy-2-buten-1-ol (Rudzinski et al., 2002; Rudzinski, 2004).

K. J. Rudzinski

268

1.0

[iso] [iso]o 0.5

0.0 0

1

2 -3

time × 10 , s

1.0

[S(IV)] [S(IV)]o 0.5

0.0

0

1

2 -3

time × 10 , s

Figure 3. Decay of S(IV) and isoprene during autoxidation of Na2SO3 at 25 oC and pH = 8.0 – 8.5, initiated by thermal decomposition of K2S2O8: 1 mM ( ), 0.2 mM ( ) and 0.1 mM ( ), or catalysed by 0.01 mM MnSO4 ( ). Atmospheric significance of isoprene transformations Heterogeneous formation of HULIS and secondary aerosols from isoprene, as well as the aqueous-phase formation of tetraols were demonstrated only in simple laboratory experiments. Similarities in spectra and chromatograms of the laboratory samples and of the real-life aerosols collected in the field campaigns strongly support the hypothesis that the heterogeneous and multiphase chemistry influence the atmospheric transformation and global balance of isoprene, as well as formation of atmospheric aerosols. The hypothesis, however, requires further proof and a quantitative assessment, which would take into account the competition from other highly reactive organic components of the atmosphere. Such goals can

Heterogeneous and Aqueous-Phase Transformations of Isoprene

269

only be achieved through concerted efforts of analytical, atmospheric and organic chemists, as well as modellers. Pedersen and Sehested (2002) showed that the aqueous-phase reaction of isoprene with ozone was insignificant for the processing of isoprene in the atmosphere. They estimated the overall and individual lifetimes of isoprene due to reactions with ozone and the hydroxyl radical, at 25 oC and typical in-cloud conditions. The results (Table 3) indicate that clouds generally should not contribute much to the processing of isoprene in the atmosphere. Only in the aqueous phase, were the lifetimes of isoprene due to reactions with ozone and with OH radicals comparable. Similar conclusions were drawn for methyl vinyl ketone, while for methacrolein the clouds could reduce the overall atmospheric lifetime by 50 %. Table 3.

Individual and overall lifetimes of isoprene in the atmosphere, due to reactions with 0.01 pptv OH radicals and 30 ppbv O3, at 25 oC (Pedersen and Sehested, 2001).

Liquid water content Lifetime (OH), h Lifetime (O3), h Lifetime (OH + O3), h

Gas-phase 0 1.12 29.4 -

10-7 28.6 1.08

Cloud 10-6 28.6 1.08

10-5 28.6 1.08

Aqueous phase 1 1 2.05 -

Rudzinski (2004) compared the rates of the aqueous-phase reaction of isoprene with sulphate radicals against the rates of the gas- and the aqueous-phase reactions of isoprene with OH radicals, NO3 radicals and ozone. The rates were evaluated for 25 oC, typical atmospheric concentrations of reactants (Herrmann et al., 2000), and a LWC of 10-5 and 10-4. The partitioning of reactants between phases was described using Henry’s Law. The results, shown in part in Table 4, indicated that the aqueous-phase reaction of isoprene with sulphate radicals was competitive against other reactions only in the aqueous phase and at very high values of liquid water content (LWC =10–4). Table 4. Reactant X

Concentration in the gas-phase

rX,g /(rSO4,aq LWC)* (LWC = 10-4)

rX,aq /rSO4,aq * aqueous phase

OH

0.04 pptv 0.0004 pptv 0.33 pptv 0.0017 pptv 90 ppbv 30 ppbv

684.0 68.4 38.3 0.19 196 65.4

1.89 0.02 0.2 0.001 0.2 0.066

NO3 O3 *

Relative rates of isoprene reactions with OH, NO3, O3 and SO4•in the gas phase (rX,g) and in the aqueous phase (rX,aq) at 25 oC.

concentration of SO4•- radicals in the aqueous phase was 10-12 M

The analysis presented here of the atmospheric significance of the reactions has ignored several important factors such as chemical nonlinearity of the reacting systems, microphysical redistribution of reactants in real atmospheric processes, the kinetic character

K. J. Rudzinski

270

of uptake processes and the influence of chemical reactions on the solubility and the uptake of reactants. For instance, chemical reactions coupled with the mass transfer processes to and in the droplets can significantly alter the Henry’s Law partitioning of reactants based on the bulk-phase concentrations. Nonlinearities of the coupled processes can induce instabilities and oscillations already observed in simple laboratory systems (Pasiuk-Bronikowska et al., 2000a, 2003). Furthermore, the coalescence and break-up of droplets can change the concentrations of reactants inferred from the mass transfer and chemical reaction considerations. Thus, the presented conclusions may be biased and should be re-examined using a more thorough modelling procedure. There are also general reasons that justify further research on the heterogeneous and the aqueous-phase transformations of isoprene. Several reviews on atmospheric processes reflect a common awareness of a general importance of heterogeneous and multiphase reactions (Fuzzi and Ebel, 2000; Herrmann, 2003; Ravishankara, 2003). As already mentioned, von Kuhlmann et al. (2003) showed by modelling that the global balance of isoprene did not close, and called for research on heterogeneous sinks for isoprene. Moreover, the heterogeneous and the aqueous-phase chemistry of isoprene may play a role in local air quality and health problems, that occur close to the sources of isoprene – in the cities, forests, offices and habitats, transport facilities and vehicles, and in the industry. For example, isoprene or products of isoprene reaction with ozone cause eye and airways irritation (Wilkins et al., 2001; Wolkoff and Nielsen, 2001, Rohr et al., 2002). Isoprene health risks in the rubber industry have also been studied (Bird et al., 2001). The possible heterogeneous and multiphase reaction sites in living organisms, seawater and soils were referenced in the Introduction section. partial gas-phase oxidation

ozonides, peroxides

organic compounds

submicron particles heat balance

emission to atmosphere

feeding

soil, water

blue haze

growth

larger particles

plants

Figure 4.

condensation and agglomeration

deposition, rain and snow

color white => grey => => brown => black

Feedbacks in a vegetation - atmosphere system described by Went (1960).

Last but not least come the environmental feedbacks that link atmospheric chemistry, the biosphere and climate change. Isoprene is a major biogenic trace component of the atmosphere. It was more than 40 years ago when Went (1960) discussed a specific relation between plant emissions and a blue haze observed above the countryside vegetation on sunny

Heterogeneous and Aqueous-Phase Transformations of Isoprene

271

days. He postulated a cycle that started with plants and the organic compounds they emitted, that were oxidised in the atmospheric gas-phase to compounds that condensed to minute particles that caused the blue haze, and also influenced the heat balance of the ecosystem (and consequently the growth of plants and the emissions). The minute particles grew or agglomerated to larger particles that deposited with rain and snow to build up and possibly fertilise the soil, which influenced plants and the emission of organic compounds (Figure 4). This was a beautiful example of feedback relations that might involve isoprene.

? light CO2

forest growth

T

terpenoid emission

?

heterogeneous and aqueousphase reactions biodiversity

leaf physiology, membranes, damage

?

Figure 5.

aerosol formation

gas-phase reactions

SO2, NOx, NOy, O3, OH, NO3 SO2 LRT, radical scavenging

LRT NOx, NOy

Atmosphere-biosphere-climate feedbacks induced by increased temperature and contents of CO2 (details and references are given in the text).

Five years later Rasmussen and Went (1965) conducted the first systematic study of bVOC (isoprene included) contents in the air over forests, meadows and prairies at several locations in the US. Now, at the turn of the century, each year brings at least one significant review on bVOC and isoprene or isoprene alone – Kesselmeier and Staud (1999), Fuentes et al. (2000), Sharkey and Yeh (2001), Monson (2002), Penuelas and Llusia (2001, 2003), Lerdau and Gray (2003). New feedbacks involving isoprene are often discussed (Shallcross and Monks, 2000; Fuentes et al., 2001; Kulmala et al., 2003, 2004). The role of isoprene in formation of aerosols was discussed in a review on atmospheric aerosols that was published almost simultaneously with the beginning of this Workshop (Kanakidou et al., 2004).

272

K. J. Rudzinski

The biosphere-atmosphere-climate feedbacks are quite complicated. A simplified scheme in Figure 5 shows what can happen in the atmosphere and biosphere when temperature and the CO2 concentration increase. Such conditions are expected to promote the growth of plants and, consequently, the emission of terpenoids (including isoprene), which are converted to a larger amount of aerosols that shade the Earth, also by cloud formation, to reduce the temperature (Kulmala et al., 2003, 2004). This relatively simple feedback action is complicated by several factors. Rises in temperature increase the emission of bVOC directly, but within limits determined by such factors as enzyme degradation and physiological response (Clark et al., 2003; Penuelas and Llusia, 2003). Thus, excessive heating may reduce the emissions. Direct influence of CO2 on bVOC emission is species specific – for instance the production of isoprene increases in some plants, but decreases in others (Loreto et al., 2001; Penuelas and Llusia, 2001). Increased production of aerosols and cloud formation or coating of vegetation with particles may limit the amount of light available to the plants and hence decrease the plant growth (Graham et al., 2003; Grantz et al., 2003). Fuentes et al. (2001) described a multilevel mechanism by which VOC and aerosols can increase the nocturnal temperatures of the atmosphere. The mechanism employed absorption of terrestrial radiation by VOC, the release of latent heat during condensation of water vapour on aerosols and thermal stabilisation of the atmosphere. On a long time scale, the mechanism would induce the successful expansion of neotropical plants, which produce more VOC (not shown in Figure 5). The gas-phase reactions of terpenoids with other trace compounds present in the atmosphere may result in formation of substances that are harmful to plants (Cape, 2003; Wannaz et al., 2003). Good condition of plant leaves is critical for emission of bVOC and photosynthesis, so degradation of plants translates into reduced growth and emissions. On the other hand, the long-range transport of nitrogen stored in such gas-phase products as PAN and MPAN may contribute to remote fertilization and growth enhancement (Shallcross and Monks, 2000; Fuentes et al., 2001). Many pollutants deposit on vegetation, and hence can enter and destroy plant tissues either directly or after transformation to even more harmful substances (Erisman and van Pul, 1997; Kesselmeier et al., 1997; Wannaz et al., 2003). Heterogeneous and aqueous-phase reactions of bVOC may produce unknown harmful substances as well, but also may protect the plants by scavenging and destroying the harmful oxidants. Actually, isoprene is considered one of the most effective antioxidants in plants (Loreto et al., 2001a; Loreto and Velikova, 2001; Affek and Yakir, 2002). Degradation and protection of plant ecosystems influences directly the level of CO2 in the atmosphere and the patterns of bVOC emissions. The scheme presented here is incomplete, and does not contain several important factors like water stress or land use (Lerdau and Gray, 2003; Penuelas and Llusia, 2003). Although limited, the scheme in Figure 5 demonstrates well the complexity and nonlinearity of atmosphere-biosphere-climate interactions. Successful modelling and quantifying of this interaction requires a careful and thorough understanding of possible feedbacks, and teams and projects still more multidisciplinary than they are presently are needed to tackle it. Condensation nuclei of such cooperations already exist within several European and global initiatives such as two ESF programs – INTROP and VOCBAS, two EU networks – ACCENT and EUROCHAMP and two global networks – IGAC and IGBP. Chamber experiments with isoprene Dodge (2000) wrote an extensive review on large and small simulation chambers. Gasphase reactions of isoprene were investigated in virtually all facilities – SAPHIR in Jülich (formerly TASK, 370 m3), EUPHORE in Valencia (200 m3), UNC (University of North

Heterogeneous and Aqueous-Phase Transformations of Isoprene

273

Carolina, 150 m3), TVA (Tennessee Valley Authority, 28 m3), CSIRO (Commonwealth Scientific and Industrial Research Organization in Australia, 20 m3), EPA (US Environment Protection Agency, 9 m3), UCR (University of Carolina at Riverside, 3.5-5.8 m3), General Motors (0.7 m3) and Wuppertal reactors (0.45 – 0.6 m3). Further research on isoprene was done by Geiger et al. (2002) in EUPHORE and Karl et al. (2004) in SAPHIR. All these studies made contributions to our understanding of the gas-phase chemistry of isoprene. Chambers were also used in the research on heterogeneous reactions of isoprene and of other atmospheric trace compounds. As already discussed, Czoschke et al. (2003) studied the formation of SOA from products of isoprene oxidation in 500 dm3 Teflon-bag chambers at UNC. Folkers et al. (2003a,b,c) studied the partitioning and influence of dicarboxylic acids on aerosol formation in Aerosol Chamber in Jülich. Shantz et al. (2003) investigated the growth of aqueous organic particles and cloud condensation nuclei in the CALSPAN chamber. Iinuma et al. (2004; paper submitted to this book) studied the reaction of Į-pinene with ozone on acidic particles in the Leipzig tent-chamber (9 m3). I am not aware of any aqueous-phase studies in chambers, with the exception of a chamber described by Monod et al. (submitted to this book), which actually is similar to the reactor we have been using for the isoprene studies (Rudzinski, 2004). Thus, the field is open for attempts to introduce clouds, fogs, rains and haze into the chambers and to investigate the multiphase and heterogeneous processes therein. Many technical ideas and solutions can be borrowed for this purpose from chemical engineering. References Affek H.P. and D. Yakir; Protection by isoprene against singlet oxygen in leaves, Plant Physiol. 129 (2002) 269– 277. Andreae M. O. and P. Merlet; Emission of trace gases and aerosols from biomass burning, Global Biogeochem. Cycles 15 (2001) 955-966. Atkinson R.; Atmospheric chemistry of VOCs and NOx, Atmos. Environ. 34 (2000) 2063-2101. Atkinson R. and J. Arey; Gas-phase tropospheric chemistry of biogenic volatile organic compounds: a review, Atmos. Environ. 37 (2003) 197-219. Atkinson R. and J. Arey; Atmospheric degradation of volatile organic compounds, Chem. Rev. 103 (2003a) 4605-4638. Baek S.-O. and R.A. Jenkins; Characterization of trace organic compounds associated with aged and diluted sidestream tobacco smoke in a controlled atmosphere-volatile organic compounds and polycyclic aromatic hydrocarbons, Atmos. Environ. 38 (2004) 6583-6599. Bird M.G., J.M. Rice and J.A. Bond; Evaluation of 1,3-butadiene, isoprene and chloroprene health risks, Chem.Biol. Interact. 135–136 (2001) 1–7. Borbon A., H. Fontaine, N. Locoge, M. Veillerot and J.C. Galloo; Developing receptor-oriented methods for non-methane hydrocarbon characterisation in urban air – Part I: source identification, Atmos. Environ. 37 (2003) 4051-4064. Borbon A., H. Fontaine, N. Locoge, M. Veillerot and J.C. Galloo; Developing receptor-oriented methods for non-methane hydrocarbon characterisation in urban air – Part II: source apportionment, Atmos. Environ. 37 (2003a) 4065-4076. Broderick B.M. and I.S. Marnane; A comparison of the C2 - C9 hydrocarbon composition of vehicle fuels and urban air in Dublin, Ireland, Atmos. Environ. 36 (2002) 975-986. Boyd D.R., D. Clarke, M.C. Cleij, J.T.G. Hamilton and G.N. Sheldrake; Bacterial biotransformation of isoprene and related dienes, Monatsh. Chem. 131 (2000) 673-685. Cape J.N.; Effects of airborne volatile organic compounds on plants, Environ. Pollut. 122 (2003) 145-157.

274

K. J. Rudzinski

Chan M.N. and C.K. Chan; Hygroscopic properties of two model humic-like substances and their mixtures with inorganics of atmospheric importance, Environ. Sci. Technol. 37 (2003) 5109-5115. Claeys M., B. Graham, G. Vas, W. Wang, R. Vermeylen, V. Pashynska, J. Cafmeyer, P. Guyon, M.O. Andreae, P. Artaxo and W. Maenhaut; Formation of secondary organic aerosols through photooxidation of isoprene, Science 303 (2004) 1173-1176. Claeys M., W. Wang, A.C. Ion, I. Kourtchev, A. Gelencser and W. Maenhaut; Formation of secondary organic aerosols from isoprene and its gas-phase oxidation products through reaction with hydrogen peroxide, Atmos. Environ. 38 (2004) 4093-4098. Clark D.A., S.C. Piper, C.D. Keeling and D.B. Clark; Tropical rain forest tree growth and atmospheric carbon dynamics linked to interannual temperature variation during 1984-2000, P. Natl. Acad. Sci. USA 100 (2003) 5852-5857. Cleveland C.C. and J.B. Yavitt; Consumption of atmospheric isoprene in soil, Geophys. Res. Lett. 24 (1997) 2379-2382. Czoschke N.M., M. Jang and R.M. Kamens; Effect of acidic seed on biogenic secondary organic aerosol growth, Atmos. Environ. 37 (2003) 4287-4299. Decesari S., M.C. Facchini, E. Matta, M. Mircea, S. Fuzzi, A.R. Chughtai and D.M. Smith; Water soluble organic compounds formed by oxidation of soot, Atmos. Environ. 36 (2002) 1827-1832. Diskin A.M., P. Spanel and D. Smith; Time variation of ammonia, acetone, isoprene and ethanol in breath: a quantitative SIFT-MS study over 30 days, Physiol. Meas. 24 (2003) 107-119. Dodge M.C.; Chemical oxidant mechanisms for air quality modeling: critical review, Atmos. Environ. 34 (2000) 2103-2130. Elkanzi E.M. and G. Bee Kheng, H2O2/UV degradation kinetics of isoprene in aqueous solution, J. Hazard. Mater. 73 (2000) 55-62. Erisman J.W. and A. van Pul; Assessment of dry deposition and total acidifying loads in Europe, in S. Slanina (ed.), Biosphere-atmosphere exchange of pollutants and trace substances (Transport and chemical transformation of pollutants in the troposphere, vol. 4), Springer Verlag, Berlin – Heidelberg (1997) 93-116. Fall R. and S.D. Copley; Bacterial sources and sinks of isoprene, a reactive atmospheric hydrocarbon, Environ. Microbiol. 2 (2000) 123-130. Fan J. and R. Zhang; Atmospheric oxidation mechanism of isoprene, Environ. Chem. 1 (2004) 140-149. doi: 10.1071/EN04045. Feng J. and D. Möller; Characterization of water-soluble macromolecular substances in cloud water, J. Atmos. Chem. 48 (2004) 217-233. Fenske J.D. and S.E. Paulson; Human breath emissions of VOCs, J. Air Waste Manage. Assoc. 49 (1999) 594598. Folkers M., Th. F. Mentel, A. Wahner, Influence of organic aerosol components on the reactivity of aqueous aerosols, Geophys. Res. Abst. 5 (2003a) 06053. Folkers M., T.F. Mentel, H. Henk, R. Tillmann, A. Wahner, R.P. Otjes, M.J. Blom and H.M. Ten Brink; Partitioning of organic aerosol components between gas phase and particulate phase, Eos. Trans. AGU 84 (2003b) Fall Meet. Suppl., Abst. A51F-0746. Folkers M., Th.F. Mentel and A. Wahner; Influence of an organic coating on the reactivity of aqueous aerosols probed by the heterogeneous hydrolysis of N2O5, Geophys. Res. Lett. 30 (2003c) 1644, doi:10.1029/ 2003GL017168. Fuentes J.D., M. Lerdau, R. Atkinson, D. Baldocchi, J.W. Bottenheim, P. Ciccioli, B. Lamb, C. Geron, L. Gu, A. Guenther, T.D. Sharkey and W. Stockwell; Biogenic hydrocarbons in the atmospheric boundary layer: a review, B. Am. Meteorol. Soc. 81 (2000) 1537 – 1575. Fuentes J.D., B.P. Hayden, M. Garstang, M. Lerdau, D. Fitzjarrald, D.D. Baldocchi, R. Monson, B. Lamb, C. Geron; New Directions: VOCs and biosphere-atmosphere feedbacks, Atmos. Environ. 35 (2001) 189-191. Fuzzi, S. and A. Ebel; Research on clouds within EUROTRAC, in: P. Borrell, P. M. Borrell (eds), Transport and Chemical Transformation of Pollutants in the Troposphere. Springer, Berlin (2000) 45–68.

Heterogeneous and Aqueous-Phase Transformations of Isoprene

275

Geiger H., I. Barnes, K.H. Becker, B. Bohn, T. Brauers, B. Donner, H.-P. Dorn, M. Elend, C. M. Freitas Dinis, D. Grossmann, H. Hass, H. Hein, A. Hoffmann, L. Hoppe, F. Hulsemann, D. Kley, B. Klotz, H. G. Libuda, T. Maurer, D. Mihelcic, G. K. Moortgat, R. Olariu, P. Neeb, D. Poppe, L. Ruppert, C.G. Sauer, O. Shestakov, H. Somnitz, W. R. Stockwell, L.P. Thuner, A. Wahner, P. Wiesen, F. Zabel, R. Zellner and C. Zetzsch; Chemical Mechanism Development: laboratory studies and model applications, J. Atmos. Chem. 42 (2002) 323-357. Gelencsér A., A. Hoffer, G. Kiss, E. Tombácz, R. Kurdi and L. Bencze; In-situ formation of light-absorbing organic matter in cloud water, J. Atmos. Chem. 45 (2003) 25-33. Graham E.A., S.S. Mulkey, K. Kitajima, N.G. Phillips and S.J. Wright; Cloud cover limits net CO2 uptake and growth of a rainforest tree during tropical rainy seasons, P. Natl. Acad. Sci. USA 100 (2003) 572-576. Grantz D.A., J.H.B. Garner and D.W. Johnson; Ecological effects of particulate matter, Environ. Int. 29 (2003) 213- 239. Guenther A., C. N. Hewitt, D. Erickson, R. Fall, C. Geron, T. Graedel, P. Harley, L. Klinger, M. Lerdau, W. A. McKay, T. Pierce, B. Scholes, R. Steinbrecher, R. Tallamarju, J. Taylor and P. Zimmerman; A global model of natural volatile organic compound emissions, J. Geophys. Res. (D) 100 (1995) 8873–8892. Harley P.C, R.K. Monson and M.T. Lerdau; Ecological and evolutionary aspects of isoprene emission from plants, Oecologia 118 (1999) 109-123. Havers N., P. Burba, J. Lambert and D. Klockow; Spectroscopic characterization of humic-like substances in airborne particulate matter; J. Atmos. Chem. 29 (1998) 45-54. Herrmann, H.; Kinetics of aqueous phase reactions relevant for atmospheric chemistry, Chem. Rev. 103 (2003) 4691 – 4716. Herrmann H., B. Ervens, H.-W. Jacobi, R. Wolke, P. Nowacki and R. Zellner; CAPRAM2.3: A chemical aqueous phase radical mechanism for tropospheric chemistry, J. Atmos. Chem. 36 (2000) 231–284. Iinuma Y., O. Böge, T. Gnauk and H. Herrmann; Aerosol-chamber study of the a-pinene/O3 reaction: influence of particle acidity on aerosol yields and products, Atmos. Environ. 38 (2004) 761-773. Jang M., N.M. Czoschke, S. Lee, and R.M. Kamens; Heterogeneous atmospheric aerosol production by acidcatalyzed particle-phase reactions, Science 298 (2002) 815-817. Kanakidou M., J.H. Seinfeld, S.N. Pandis, I. Barnes, F.J. Dentener, M.C. Facchini, R. van Dingenen, B. Ervens, A. Nenes, C.J. Nielsen, E. Swietlicki, J.P. Putaud, Y. Balkanski, S. Fuzzi, J. Horth, G.K. Moortgat, R. Winterhalter, C.E.L. Myhre, K. Tsigaridis, E. Vignati, E.G. Stephanou and J. Wilson; Organic aerosol and global climate modelling: a review, Atmos. Chem. Phys. Discuss. 4 (2004) 5855-6024. Karl M., Th. Brauers, H.-P. Dorn, F. Holland, M. Komenda, D. Poppe, F. Rohrer, L. Rupp, A. Schaub and A. Wahner, Kinetic study of the OH-isoprene and O3-isoprene reaction in the atmosphere simulation chamber, SAPHIR, Geophys. Res. Lett. 31 (2004), L05117, doi:10.1029/2003GL019189. Karl T., P. Prazeller, D. Mayr, A. Jordan, J. Rieder, R. Fall and W. Lindinger; Human breath isoprene and its relation to blood cholesterol levels: new measurements and modeling, J. Appl. Physiol. 91 (2001) 762-770. Kesselmeier J., K. Bode, J.K. Schjørring and R. Conrad; Biological mechanisms involved in the exchange of trace gases, in S. Slanina (ed.), Biosphere-atmosphere exchange of pollutants and trace substances (Transport and chemical transformation of pollutants in the troposphere, vol. 4), Springer Verlag, Berlin – Heidelberg (1997) 117-134. Kesselmeier J. and M. Staudt; Biogenic Volatile Organic Compounds (VOC): An Overview on Emission, Physiology and Ecology, J. Atmos. Chem. 33 (1999) 23-88. Kiss G., E. Tombacz, B. Varga, T. Alsberg and L. Persson; Estimation of the average molecular weight of humic-like substances isolated from fine atmospheric aerosol, Atmos. Environ. 37 (2003) 3783-3794. Kulmala M., T. Suni, K.E.J. Lehtinen, M. Dal Maso, M. Boy, A. Reissell, U. Rannik, P. Aalto, P. Keronen, H. Hakola, J.Baeck, T. Hoffmann, T. Vesala and P. Hari; A new feedback mechanism linking forests, aerosols, and climate, Atmos. Chem. Phys. 4 (2004) 557-562 SRef-ID: 1680-7324/acp/2004-4-557. Leber A.P.; Overview of isoprene monomer and polyisoprene production processes, Chem.-Biol. Interact. 135136 (2001) 169-173. Lerdau M. and D. Gray; Ecology and evolution of light-dependent and light-independent phytogenic volatile organic carbon, New Phytol. 157 (2003) 199-211.

276

K. J. Rudzinski

Limbeck A., M. Kulmala and H. Puxbaum; Secondary organic aerosol formation in the atmosphere via heterogeneous reaction of gaseous isoprene on acidic particles, Geophys. Res. Lett., 30 (2003) 1996 doi: 10.1029/2003GL017738 09. Logan B.A., R.K. Monson and M.J. Potosnak; Biochemistry and physiology of foliar isoprene production, Trends Plant Sci. 5 (2000) 477-481. Loreto F., R.J. Fischbach, J.-P. Schnitzler, P. Ciccioli, E. Brancaleoni, C. Calfapietra and G. Seufert; Monoterpene emission and monoterpene synthase activities in the Mediterranean evergreen oak Quercus ilex L. grown at elevated CO2 concentrations, Global Change Biol. 7 (2001) 709-717. Loreto F., M. Mannozzi, C. Maris, P. Nascetti, F. Ferranti, and S. Pasqualini; Ozone quenching properties of isoprene and its antioxidant role in leaves, Plant Physiol. 126 (2001a) 993-1000. Loreto F. and V. Velikova; Isoprene produced by leaves protects the photosynthetic apparatus against ozone damage, quenches ozone products, and reduces lipid peroxidation of cellular membranes, Plant Physiol. 127 (2001) 1781-1787. Matsunaga S.. M. Mochida, T. Saito and K. Kawamura; In situ measurement of isoprene in the marine air and surface seawater from the western North Pacific, Atmos. Environ. 36 (2002) 6051-6057. Melnick R.L. and R.C. Sills; Comparative carcinogenicity of 1,3-butadiene, isoprene, and chloroprene in rats and mice, Chem.-Biol. Interact. 135-136 (2001) 27-42. Merck; Safety Data Sheets 2-metyl-1,3-butadiene (stabilised) for synthesis, Merck Schuchardt OHG, Hochenbrunn (1999). Milne, P. J., D. D. Riemer, R. G. Zika and L. E. Brand; Measurement of vertical distribution of isoprene in surface seawater, its chemical fate, and its emission from several phytoplankton monocultures, Mar. Chem. 48 (1995) 237–244. Monson R.K.; Volatile organic compound emissions from terrestrial ecosystems: a primary biological control over atmospheric chemistry, Israel J. Chem. 42 (2002) 29-42. NIST Chemistry WebBook; http://webbook.nist.gov/chemistry, (search for isoprene) (2001), accessed on 24th Jan. 2003. Pasiuk-Bronikowska W., T. Bronikowski, K.J. Rudzinski and J. Ziajka; Transformations of atmospheric constituents and pollutants induced by S(IV) autoxidation – chemistry and kinetics, in: U. Schurath, J. Peeters, H. Herrmann, D. Poppe (eds), Chemical mechanism development. EUROTRAC-2 CMD Annual Report 2000, EUROTRAC-2 ISS, Munich (2000) 123-126. Pasiuk-Bronikowska W., T. Bronikowski and M. Ulejczyk; Study of the kinetics of S(IV) autoxidation inhibited by phenolics, in: M.J. Rossi, E.-M. Rossi (eds), Chemical Processes and Mechanisms - Chemical Mechanism Development. Proc. EC/EUROTRAC-2 Joint Workshop, EPFL, Lausanne (2000) 144-147. Pasiuk-Bronikowska W., T. Bronikowski and M. Ulejczyk; Inhibition of the S(IV) autoxidation in the atmosphere by secondary terpenic compounds, J. Atmos. Chem. 44 (2003) 97-111. Pedersen T. and K. Sehested; Rate constants and activation energies for ozonolysis of isoprene, methacrolein and methyl-vinyl-ketone in aqueous solution: significance to the in-cloud ozonation of isoprene, Int. J. Chem. Kin. 33 (2001) 182-190. Penuelas J. and J. Llusia; The complexity of factors driving volatile organic compound emissions by plants, Biol. Plantarum 44 (2001) 481-487. Penuelas J. and J. Llusia; BVOCs: plant defence against climate warming?, Trends Plant Sci. 8 (2003) 105-109. Pouli A.E., D. G. Hatzinikolaou, C. Piperi, A. Stavridou, M. C. Psallidopoulos and J.C. Stavrides; The cytotoxic effect of volatile organic compounds of the gas phase of cigarette smoke on lung epithelial cells, Free Rad. Biol. Med. 34 (2003) 345-355. Rasmussen R.A. and F.W. Went; Volatile organic material of plant origin in the atmosphere, P. Natl. Acad. Sci. USA 53 (1965) 215-220. Ravishankara A.R.; Introduction: Atmospheric Chemistry-Long-Term Issues, Chem. Rev. 103 (2003) 4505 – 4508.

Heterogeneous and Aqueous-Phase Transformations of Isoprene

277

Rodriguez-Concepcion M. and A. Boronat; Elucidation of the methylerythritol phosphate pathway for isoprenoid biosynthesis in bacteria and plastids. A metabolic milestone achieved through genomics, Plant Physiol. 130 (2002) 1079-1089. Rohr A.C., C.K. Wilkins, P.A. Clausen, M. Hammer, G.D. Nielsen, P. Wolkoff and J.D. Spengler; Upper airway and pulmonary effects of oxidation products of (+)-Į-pinene, d-limonene, and isoprene in balb/c mice, Inhal. Toxicol. 14 (2002) 663-684. Rudzinski K.J.; Degradation of isoprene in the presence of sulphoxy radical anions, J. Atmos. Chem. 48 (2004) 191–216. Rudzinski K.J. and W. Pasiuk-Bronikowska; Isoprene inhibition of S(IV) autoxidation initiated by peroxydisulphate, in: P.M. Midgley, M. Reuther, M. Williams (eds), Transport and chemical transformation in the troposphere. Proc. EUROTRAC-2 Symposium 2000, Springer Verlag, Berlin (2001) 1-4 (CD). Rudzinski K.J., W. Pasiuk-Bronikowska and J. Królik; Mechanistic study of isoprene inhibition of S(IV) autoxidation in aqueous phase, in: M.J. Rossi, E.-M. Rossi (eds), Chemical Processes and Mechanisms Chemical Mechanism Development. Proc. EC/EUROTRAC-2 Joint Workshop, EPFL, Lausanne (2000) 148-151. Rudzinski K.J., W. Pasiuk-Bronikowska and J. Królik; Chemical interactions of precursors of tropospheric aerosols – SO2 and isoprene, in: P.M. Midgley, M. Reuther (eds), Transport and chemical transformation in the troposphere. Proc. of EUROTRAC-2 Symposium 2002, Margraf Verlag, Weikersheim and Backhuys Publishers, Leiden (2002), CMD-13, 1-5. Saunders S.M., M.E. Jenkin, R.G. Derwent and M.J. Pilling; Protocol for the development of the Master Chemical Mechanism, MCM v3 (Part A): tropospheric degradation of non-aromatic volatile organic compounds, Atmos. Chem. Phys. 3 (2003) 161-180. Shallcross D.E. and P.S. Monks; New directions: A role for isoprene in biosphere-climate-chemistry feedbacks, Atmos. Environ. 34 (2000) 1659-1660. Shantz, N.C., W.R. Leaitch and P.F. Caffrey; Effect of organics of low solubility on the growth rate of cloud droplets, J. Geophys. Res. 108 D5 (2003) 4168 doi: 10.1029/2002JD002540. Sharkey T.D. and S. Yeh; Isoprene emission from plants, Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 (2001) 407-436. von Kuhlmann R., M. G. Lawrence, U. Pöschl and P. J. Crutzen; Sensitivities in global scale modeling of isoprene, Atmos. Chem. Phys. 4 (2004) 1-17 SRef-ID: 1680-7324/acp/2004 4 1. Wagner W.P., M. Nemecek-Marshall and R. Fall; Three distinct phases of isoprene formation during growth and sporulation of Bacillus subtilis, J. Bacteriol. 181 (1999) 4700-4703. Wannaz E.D., J.A. Zygadlo and M.L. Pignata; Air pollutants effect on monoterpenes composition and foliar chemical parameters in Schinus areira L., Sci. Total Environ. 305 (2003) 177-193. Went F.W.; Blue hazes in the atmosphere, Nature 187 (1960) 641-643. Wilkins K. and K. Larsen; Volatile organic compounds from garden waste, Chemosphere 32 (1996) 32, (10), 2049-2055. Wilkins C.K., P.A. Clausen, P. Wolkoff, S.T. Larsen, M. Hammer, K. Larsen, V. Hansen and G.D. Nielsen; Formation of strong airway irritants in mixtures of isoprene/ozone and isoprene/ozone/nitrogen dioxide, Environ. Health Persp. 109 (2001) 937-942. Wolkoff P. and G.D. Nielsen; Organic compounds in indoor air – their relevance for perceived indoor air quality? Atmos. Environ. 35 (2001) 4407-4417.

Investigation of Atmospheric Transformations of Diesel Emissions in the European Photoreactor (EUPHORE) Barbara Zielinska1, John Sagebiel1, William Stockwell1, Jake McDonald2, JeanClare Seagrave2, Peter Wiesen3, and Klaus Wirtz 4 1

Desert Research Institute, 2215 Raggio Parkway, Reno, Nevada 89512, USA; 2Lovelace Respiratory Research Institute, Albuquerque, NM, USA; 3University of Wuppertal, Germany; 4Fundacion Centro de Estudios Ambientales del Mediterraneo, Valencia, Spain Key Words: Diesel exhaust, Emissions, Particles, Photoreator, PAHs, Toxicity

Introduction Once released into the atmosphere, primary diesel emissions (or any other direct emissions) are subject to dispersion and transport and, at the same time, to various physical and chemical processes, which determine their ultimate environmental fate. To elucidate potential health effects of diesel exhaust, it is insufficient to characterize only the primary pollutants that are directly emitted from diesel vehicles. Secondary compounds, formed during the transport of emissions through the atmosphere, may also affect human health. As primary diesel emissions are a very complex mixture containing thousands of organic and inorganic constituents in the gas and particulate phases that have different chemical reactivities, the rates and mechanisms of their atmospheric transformation processes ultimately determine the atmospheric lifetimes of the initially emitted species as well as the chemical nature and biological activities of the resulting secondary pollutants. The more reactive compounds with short lifetimes will be removed from the atmosphere relatively quickly, whereas more stable pollutants (including atmospheric transformation products) can be transported over greater distances. Some of the gaseous species, by a series of chemical transformations, are converted into particles, forming secondary aerosols. Sulfates and nitrates are the most common secondary particles, though a fraction of particulate organic carbon also results from volatile organic compounds (VOC) via atmospheric reactions with reactive gaseous species such as HO radicals, NO3 radicals, or O3. Clearly, knowledge of the atmospheric loss processes, transformation products, and lifetimes for automotive emissions is important, because these factors determine the biological activities and geographic extent of the influence of these emissions. This paper describes a project that was designed to study the products from sunlight, ozone, HO radical, and NO3 radical-initiated reactions (the latest in the dark) of diesel emissions with the aid of an environmental simulation chamber, under realistic ambient conditions (dilution in the range of 1:300 – 1: 400). The European Photoreactor in Valencia, Spain, (EUPHORE) which is currently one of the largest (approximately 200 m3) and best equipped outdoor simulation chambers in the world, is employed for this study. Experimental The EUPHORE chambers in Valencia, Spain, are described in detail elsewhere (Becker, 1996). Briefly, there are two chambers, each consisting of a half-spherical Teflon bag with a volume of about 204 m3 (see Figure 1). The chambers are protected against atmospheric influences by two half spherically shaped protective housings. The inlet and 279 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 279–284. © 2006 Springer. Printed in the Netherlands.

B. Zielinska et al.

280

outlet ports and other accessories, such as mixing fans, analytical systems, and mechanical excess pressure valves, are located on the floor so that the chamber spherical surface remains unobstructed. Integrated into the flanges are ports for the input of reactants and the sampling lines for different analytical instruments. To mix the reactants, two mixing fans with an air throughput of 4000 m3/h each are installed in each chamber. White mirror systems for in situ FTIR spectrometer measurements are situated in the chamber. Numerous analytical instruments, such as NO/NOx/NO2 analyzers, GC, HPLC, GC-MS-system, SMPS and an ozone analyzer are mounted on a platform under the chamber floor. The sampling lines for these instruments are connected directly to the chamber.

Figure 1.

The European Photoreactor EUPHORE in Valencia/Spain.

Diesel exhaust is generated on-site using a light-duty diesel engine and dynamometer, which is equipped with a Horiba gas analyzer allowing for the continuous monitoring of engine-out gaseous emissions of CO, CO2, NOx, total hydrocarbons (THC) and O2. A modern diesel engine (i.e. with a common rail direct injection, turbocharged, intercooled engine) was obtained from Ford Motor Company (Lynx V277 90 PS Ford Focus engine) and has been mounted on the dynamometer. The engine test rig is situated below the EUPHORE chamber; Figure 2 shows the position of the engine in relation to the chamber. The exhaust is guided in heated tubes under the chamber and is introduced into the chamber by means of a three-way valve, which can tolerate temperatures up to 200oC. If desired, a separate, smaller line can be used that allows approximately 20% of the total exhaust from the engine into the chamber, the rest is vented. Inside the EUPHORE chamber very fast dilution is obtained by mixing the exhaust gas with the clean air with help of a powerful fan. This prevents the agglomeration of particles and the condensation of water. The diesel engine is operated under a load of 60 Nm (Newton-meters) that represents approximately 50% of total engine power. Prior to introduction of exhaust to the chamber, the engine is warmed up for approximately 30 min to reach steady-state conditions. A diesel fuel that most closely matches what is currently used (or will be used in the near future) in most of

Investigation of Atmospheric Transformations of Diesel Emissions

281

the United States is employed (47 ppm sulfur and 15% aromatic content). The exhaust is introduced to the chamber over approximately 2 min and 6 min to obtain sufficient material for chemical analyses and toxicity testing, respectively.

Figure 2.

Set-up of the engine test rig at the EUPHORE chamber.

Results and Discussion The test matrix that will be carried out in the EUPHORE chamber is shown in Tables 1 and 2. These series of tests provide experiments that will examine the effects of aging, photolysis, HO, O3 and the NO3 radical (in the dark) on the composition of diesel exhaust. The experiments are divided into dark (Table 1) and light (Table 2) exposures test matrixes. The dark experiments D-1, D-2, and D-5 provide baselines for other experiments. Experiment D-3 investigates the effects of O3 on diesel exhaust in the dark and experiment D-4 allows the effect of the NO3 radical on diesel exhaust to be studied. Dinitrogen pentoxide is used as a source of NO3 radicals. N2O5 is prepared by reacting ozone with NO2 directly in the chamber. Toxicity testing requires relatively large samples (on the order of 50 mg). Therefore, experiments D-5 through D-7 are made to duplicate the first experiments but without significant amounts of sample extraction for chemical analyses, although other continuous monitoring is performed. The light exposure test matrix includes six core experiments. L-1 is a baseline experiment for the light exposure series; it examines the effects of photolysis reactions on the exhaust composition. Experiment L-2 examines the effect of HO reactions on exhaust composition at relatively high NOx. Experiment L-3 is proposed to examine the effect of the

B. Zielinska et al.

282

reactions of HO on exhaust composition at lower NOx by addition of H2O2. Hydrogen peroxide photolyses to produce HO but no additional NOx will be added to the chamber when H2O2 is added. Toxicity testing is also proposed for the light experiments. Therefore experiments L-4 through L-6 are made to duplicate the first experiments but without significant amounts of sample extraction for chemical analyses, although other continuous monitoring is performed. Chamber contents will be collected for toxicity testing at the end of experiments L-4 through L-6. Table 1.

Dark Exposure Test Matrix.

Run D-1

Composition Diesel Exhaust Only

D-2 D-3

HONO + Diesel Exhaust O3 + Diesel Exhaust

D-4

N2O5 + Diesel Exhaust

D-5

Diesel Exhaust Only

D-6

O3 + Diesel Exhaust

D-7

N2O5 + Diesel Exhaust

Purpose Determine changes in exhaust composition due to aging in chamber. Provide baseline for HO experiment, L-2. Study reactions of ozone with diesel exhaust in the dark. Investigate effects of NO3 on diesel exhaust composition. N2O5 decomposes to form NO2 and NO3. Duplicate run of D-1 for collection of samples for toxicity testing. Duplicate run of D-3 for collection of samples for toxicity testing. Duplicate run of D-4 for collection of samples for toxicity testing.

The initial series of dark experiments was carried out in January 2005 in the EUPHORE chamber. However, it became apparent during these experiments that our modern diesel engine equipped with an oxidation catalyst had a very low diesel particulate matter (PM) emission rate (especially with the low sulfur, low aromatic fuel, we used), and relatively high NOx emissions under load. In order to obtain the required diesel PM concentrations in the chamber, a longer delivery time of the raw exhaust was necessary (on the order of 2 to 6 minutes). This resulted in unrealistically high concentrations of NOx in the chamber. For example, 2-min introduction of diesel exhaust to the chamber produced approximately 30 µg/m3 of diesel PM and nearly 1 ppm of NOx (30% of NOx as NO2). Because of the high NOx concentrations in the chamber, it was not possible to carry out certain exposure scenarios, for example dark ozone exposures. Thus, we are in the process of developing a new NOx denuder that will be able to decrease significantly high NOx concentrations and are planning to deploy it for the next EUPHORE series of experiments, planned for May 2005. The denuder uses cobalt oxide as an efficient sorption material for the capture of nitrogen oxides (NO, NO2, and HNO3) from exhaust streams (Braman et al., 1986; Ammann 2001; Arens et al., 2001). This denuder will need to be scaled to handle the concentration (up to 400 ppm) and flow rate (180 liters per minute) of the exhaust flow. The experiments carried out during the January 2005 EUPHORE campaign consisted mostly of D-1, D4, D-5 and D-7 runs (Table 1). To determine the particle size, number and volume concentrations the Scanning Mobility Particle Analyzer (SMPS) was used and the particle mass was continuously monitored using TEOM. NOx and NOy species were monitored using chemiluminescence and Fourier Transfer Infrared (FTIR) instruments. After several hours of exposure, the samples for semivolatile organic compound’s (SVOC) analysis

Investigation of Atmospheric Transformations of Diesel Emissions

283

were collected using a polystyrene-divinylbenzene resin (XAD)-coated annular denuder (Gundel et al., 1995) followed by a 90 mm Teflon-impregnated glass fiber (TIGF) filter and an XAD cartridge. The flow rate of 100 L/min was used. The samples for toxicity evaluation were collected using 8x10-inch Teflon filters followed by two XAD cartridges, with flow rate of 200 L/min. In addition, canister samples were collected for volatile organic compounds (VOC) analysis; dinitrophenylhydrazine (DNPH) impregnated cartridges for carbonyl compound analysis, and diesel particles on quartz fiber filters for organic and elemental carbon (OC/EC) analysis. Table 2.

Light Exposure Test Matrix.

Run Composition L-1 Diesel Exhaust Only L-2

L-3

HONO + NOx + Diesel Exhaust H2O2 + Diesel Exhaust

L-4

Diesel Exhaust Only

L-5

HONO + NOx + Diesel Exhaust H2O2 + Diesel Exhaust

L-6

Purpose Examine effects of photolysis reactions on exhaust composition. Investigate effects of HO on diesel exhaust under conditions of high NOx. Investigate effects of HO on diesel exhaust under conditions of lower NOx. Duplicate run of L-1 for collection of samples for toxicity testing. Duplicate run of L-2 for collection of samples for toxicity testing. Duplicate run of L-3 for collection of samples for toxicity testing.

The SVOC collected on each denuder-filter- XAD cartridge sampling train are currently being extracted and analyzed by high resolution capillary gas chromatography-mass spectrometry (GC/MS), as described previously (Zielinska et al., 2004). The samples are analyzed for polycyclic aromatic hydrocarbons (PAH), oxy-PAH, nitro-PAH, organic acids and diacids and other polar organics (potential transformation products), using a Varian 1200 triple quadrupole GC/MS/MS. Samples collected for toxicity testing are extracted by sonication/agitation (filters) and accelerated solvent extraction (XAD) in acetone. The toxicity of combined PM + SVOC extracts will be ranked by comparison of dose-response curves for several endpoints achieved by either in-vitro (macrophages/Salmonella) or in-vivo (instillation in rat lung) assessments broadly classified as: Genotoxicity; Cytotoxicity; Inflammation; Structural/Functional and Oxidative Stress. The results of these analyses will be reported in subsequent papers. Acknowledgement Financial support for this project is provided by the Health Effects Institute, USA. References Ammann, M., (2001): Using 13N as Tracer in Heterogeneous Atmospheric Chemistry Experiments. Radiochim. Acta, 89, 831-838 . Arens, F., L. Gutzwiller, U. Baltensperger, H. W. Gaggeler and M. Amman, (2001):Heterogeneous Reaction of NO2 on Diesel Soot Particles. Environ. Sci. Technol., 35, 2191-2199. Becker, K.H. (1996). Design and Technical Development of the European Photoreactor and First Experimental Results, Final Report of the EC-Project (EV5V-CT92-0059), Wuppertal, 1996.

284

B. Zielinska et al.

Braman, R.S. and M.A. de la Centera, (1986): Sequential, Selective Hollow Tube Preconcentration and Chemiluminescence Analysis System for Nitrogen Oxide Compounds in Air, Anal. Chem., 58, 15371541. Gundel, L.A., V.C. Lee, K.R.R. Mahanama, R.K. Stevens, and J.M. Daisey (1995). Direct Determination of the Phase Distributions of Semi-Volatile Polycyclic Aromatic Hydrocarbons Using Annular Denuders. Atmos. Environ., 29, 1719-1995. Zielinska, B., J. Sagebiel, W.P. Arnott, C.F. Rogers, K.E. Kelly, D.A. Wagner, J.S. Lightly, A.F. Sarofim, G. Palmer (2004). Phase and Size Distribution of Polycyclic Aromatic Hydrocarbons in Diesel and Gasoline Vehicle Emissions. Environ. Sci. Technol. 38, 2557-2567.

Investigation of Real Car Exhaust in Environmental Simulation Chambers: Results from the INFORMATEX and DIFUSO Projects Peter Wiesen Bergische Universität Wuppertal, Phys. Chemie / Fachbereich C, D-42097 Wuppertal, Germany Key Words: Car Exhaust, Diesel, Gasoline, Modelling, Ozone, Particles, Photoreactors

Introduction Emissions from the combustion of fossil fuels, especially road traffic emissions, are major contributors to air pollution. Though exhaust-cleaning techniques for vehicles have been significantly improved during the last years and the averaged fuel consumption per single vehicle has significantly decreased, the contribution of road traffic emissions to tropospheric photosmog formation is still high, due to the increasing total number of vehicles. For the future, an increase in the number of diesel engines is anticipated. Since this engine type consumes less fuel than a comparable gasoline engine, the replacement of gasoline-powered vehicles by diesel-fuelled cars would reduce CO2 emissions and therefore global warming and save natural resources. In contrast, possible negative influences of the relatively high number of particles emitted by diesel-powered vehicles in comparison to gasoline-fuelled cars on tropospheric chemistry is a matter of grave health and environmental concern. At present, the effect fuel composition on the related change of a long-term exchange of the fuel type from gasoline to diesel on air quality is not clear. It is well known that gasoline exhaust is characterised by relatively high concentrations of volatile organic compounds (VOC) and low NOx concentrations, whereas the VOC/NOx ratio of diesel exhaust is extremely low, due to very low concentrations of unburned fuel (Marshall and Owen, 1995, Balek et al., 1997). An effective replacement of gasoline by diesel as the major fuel used in road traffic might therefore significantly influence the photochemistry of the troposphere, e.g. ozone formation. Atmospheric simulation chambers offer the unique possibility for studying the impact of real car exhaust under almost real-world conditions on tropospheric photosmog formation. Within two EC projects, namely INFORMATEX (Becker, 1998) and DIFUSO (Wiesen, 2000) the ozone formation generated by the photooxidation of the exhaust of a gasoline- and diesel-powered engine has been investigated for different fuel formulations. The findings of these projects are summarised in the present paper with focus on the DIFUSO project. More detailed information can be found in recent papers by Geiger et al. (2002 and 2003). Eperimental The experiments presented here were carried out in the outdoor simulation chamber EUPHORE in Valencia, Spain. This facility and the analytical equipment used for the measurement of selected trace gases are described in detail by Becker (1996) as also Barnes and Wenger (1998). 285 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 285–294. © 2006 Springer. Printed in the Netherlands.

P. Wiesen

286

Within the measurement campaign, several smog experiments were carried out. Two reference experiments (base mixture I and II) were focused on a VOC "only" mixture containing n-butane, ethene and toluene in the presence of NO, representing the polluted troposphere. Both experiments differed in the initial NO/NO2 ratio, while the total amount of initial NOx was more or less similar (about 200 ppb). For each of the 5 diesel fuels investigated, a smog experiment was performed, where diesel exhaust was injected into the chamber up to a constant NOx level. After exhaust injection, about 1 ppm of the VOC mixture was added and the chamber was opened for irradiation. Diesel exhaust was generated by a diesel engine (1.8 l, 44kW) mounted on a test bed including an eddy current brake. For all smog chamber runs, similar test cycle conditions were used (see Wiesen (2000) for details). Five diesel fuels with different formulations were tested, a commercial diesel (“standard” diesel) available at European gas stations, biodiesel (rape oil methyl ester) and three sulphur-reduced diesel fuels containing defined amounts of aromatic hydrocarbons (see Table 1). Table 1.

Contents of aromatic hydrocarbons and sulphur for the 5 Diesel fuels used in the present study (Wiesen, 2000).

Fuel

Aromatic HC (weight%)

Sulphur (weight%)

29.3

0.0425

0

0.0030

3 (< 5% aromatic HC)

4.3

0.0043

4 (< 15% aromatic HC)

14.2

0.0045

5 (< 25% aromatic HC)

24.4

0.0046

1 (standard Diesel) 2 (Biodiesel)

Detailed information about the analytical equipment applied is given in Wiesen (2000). All studies were performed at atmospheric pressure. A dilution factor was obtained from the decay of inert SF6, added to the reaction mixtures as a chamber leak dilution tracer. Computer simulations The chemical simulations of the smog systems were performed using a simple chemical box model based on the Regional Atmospheric Chemistry Mechanism (RACM) by Stockwell et al. (1997). This condensed reaction scheme includes a fairly complete set of reactions for inorganic chemistry. Organic species are aggregated according to their chemical structure and reactivity into 32 stable organic species and 24 organic intermediates, except methane, ethane, ethene and isoprene, whose chemistry is treated explicitly. In total, the mechanism comprises of 237 reactions including 23 photochemical reactions. Each experiment was modelled using the box model described below, except that for diesel fuel no. 5. All computer simulations and sensitivity analyses were carried out using the box model SBOX (Seefeld, 1997, Seefeld and Stockwell 1999). This FORTRAN program incorporating the Gear algorithm (Gear, 1971) was operated on an SGI OCTANE workstation (Silicon Graphics) running under IRIX 6.5. The program uses the public domain library VODE (Brown et al., 1989) to integrate the ordinary differential equations.

Investigation of Real Car Exhaust in Environmental Simulation Chambers

287

From the measured spectral radiant meter data, photolysis frequencies, j, were calculated for several species (Wiesen, 2000). Since the box model SBOX requires photolysis frequencies in a fixed table format, the j values obtained from the radiant meter data were transformed into a suitable matrix. For the calculation of the photolysis frequencies of the 9 photosensitive “lumped” species of the RACM scheme, the program “photoRACM” (Seefeld, 1997, Seefeld and Stockwell 1999) was used. This program generates photolysis frequencies according to Madronich (1987). Since the experiments were carried out using simple VOC mixtures containing only nbutane, ethene and toluene, a few modifications of the original RACM code were necessary for better performance of the calculations. n-Butane is grouped in RACM into the surrogate species HC3, which represents the class of “low” reactive alkanes, whereas toluene is grouped into the aromatic class TOL, indicating aromatic compounds with OH reactivities equal or lower than toluene. For the present calculations, the rate coefficients for the reactions of OH radicals with HC3 (2.20 u 10-12 cm3 s-1) and TOL (5.96 u 10-12 cm3 s-1) were replaced by the explicit literature values for OH + n-butane (2.2 u 10-12 cm3 s-1 (Atkinson, 1994)) and toluene (6.6 u 10-12 cm3 s-1 (Becker, 1994)), respectively. No modifications were necessary with respect to ethene, since this hydrocarbon is explicitly treated in RACM. Results and discussion General observations The diesel exhaust added to the chamber in each experiment mainly contained CO, NOx, HCHO and higher aldehydes, SO2 and soot. The fraction of hydrocarbons, e.g. from unburned fuel, was negligible (Wiesen et al., 2000). In all experiments performed, significant formation of ozone was observed. Table 2 compares for all runs both the ozone formation rates and the values for d([O3]-[NO])/dt, which provides a measure for the ozone formation potential, derived from the experimental data 2 h after opening the chamber. Both rates are significantly smaller for the base mixture experiments than for the exhaust runs. Table 2. Comparison of d[O 3]/dt and d([O 3]-[NO])/dt derived from the experimental data 2 h after opening of the smog chamber. experiment Base Mixture I Base Mixture II Standard Diesel (fuel 1) Biodiesel (fuel 2) Diesel containing 5% Aromatic HC (fuel 3) Diesel containing 15% Aromatic HC (fuel 4) Diesel containing 25% Aromatic HC (fuel 5)

d[O3]/dt (ppb h-1) 6.2 13.1 66.9 82.7 48.8 52.8 46.0

d([O3]-[NO])/dt (ppb h-1) 48.5 46.2 116.6 106.9 108.2 58.1 74.3

The mean OH concentration, which can be estimated from the experimental 6 -3 hydrocarbon degradation, was in the range (1 –3) u 10 cm for all chamber runs performed. All exhaust/VOC studies have been compared to the base mixture I experiment, since they were carried out at an initial NO/NO2 ratio very similar to the first reference run. The ozone

288

P. Wiesen

formation rates for all other runs with diesel exhaust added were in the range 46 – 83 ppb h -1, which is a factor of about 7 –13 higher than observed in the base mixture I experiment. Due to significant deviations in the initial conditions of each experiment, such as concentrations of NO, NO2, aldehydes, etc. as well as different photolytic parameters caused by the weather conditions, a direct interpretation of the ozone data was not possible. Therefore, in order to analyse the data the smog chamber experiments were simulated using the chemical box model described above. At first it was checked whether the differences between the reference system and the exhaust system could generally account for the significant increase in ozone formation observed for the exhaust/VOC experiments. The exhaust runs were characterised by the presence of particles (soot) and relatively high initial concentrations of nitrous acid, formaldehyde and higher aldehydes. It is expected that the initial concentrations of HONO and aldehydes lead to higher ozone concentrations in the exhaust experiments, due to fast photolysis and generation of radicals. In order to check this possibility, a series of simulations with the initial concentrations of base mixture I plus several initial concentrations of nitrous acid or formaldehyde was carried out and the peak ozone concentration was calculated. The results exhibit a strong increase of peak ozone with increasing initial concentrations of HONO or HCHO, which supports the expectation that the strong ozone formation in the exhaust/VOC experiments is due to photolysis of initial nitrous acid and formaldehyde. This is also confirmed by modelling the ozone profiles of the exhaust experiments. As an example, Figure 1 illustrates the experimental ozone concentration for the standard diesel run together with calculated data.

Figure 1.

Experimental (‹) and simulated ozone profiles for different initial concentrations of nitrous acid and formaldehyde for the smog chamber run using fuel 1.

For either [HONO]0 = 0 or [HCHO]0 = 0, the calculations yield significantly lower ozone concentrations than measured. If both initial values are set to zero, the ozone formation is decreased by a factor of about 10.

Investigation of Real Car Exhaust in Environmental Simulation Chambers

289

Chemical modelling of the experiments For all runs modelled, a systematic overestimation of NO2 by the model was observed, even when the calculated concentrations of other species were excellently fitted to the measured data. The disparity occurred only for reaction times t 2 hours and further increased with increasing time. Since the measured NO2 data were corrected with respect to PAN and HNO3 interferences (Wiesen, 2000), this observation must be explained by a general lack of mechanistic information about the tropospheric degradation of toluene. It is well known that the OH-initiated degradation of aromatic hydrocarbons in smog chambers leads to a significant loss of NOx in the reaction system (see e.g. Calvert et al., 2000), which at present cannot be explained by current chemical models. Also the RACM mechanism used for the present investigations does not account for this NOx loss. Accordingly, the simulation of a reaction system including significant amounts of toluene must lead to an overestimation of NOx, as observed in each calculation performed in the present study. In the following sections the results of the chemical modelling of selected smog chamber runs are presented. Base mixture I Figure 2 illustrates the comparison of experimental and calculated concentration-time profiles for VOC, HCHO, NO, NO2 and ozone for the experiment performed with the base VOC mixture in the presence of NOx (NO/NO2 ratio = 8.8). All data (except NO2, see above) are in excellent agreement within the experimental errors, which are in the range of ±5% – ±15% for the different species. The good modelling results indicate that the RACM mechanism describes the system with sufficient accuracy.

Figure 2.

Comparison of simulated () and experimental concentrations of (left hand side) n-butane (●), ethene (…), toluene u 3 (‹),and (right hand side) NO (●), NO 2 (…), ozone (c) and formaldehyde (▲) for the smog chamber run with base mixture I.

Standard diesel (fuel 1) Figure 3 shows a comparison of experimental and calculated concentration-time profiles for NO, NO2 and ozone for the experiment performed with exhaust from the standard diesel fuel together with the base VOC mixture in the presence of NOx. Only the initial formaldehyde concentration was measured. Time-resolved data for HCHO were not available

P. Wiesen

290

for this run. Calculated and experimental concentration-time profiles are in good agreement and the system is well described by the model.

Figure 3.

Comparison of simulated () and experimental concentrations of NO (‹), NO2 (…) and ozone (c) for the smog chamber run with fuel 1.

Biodiesel (fuel 2) Figure 4 shows a comparison of experimental and calculated concentration-time profiles for HCHO, NO, NO2 and ozone for the experiment performed with exhaust from the biodiesel fuel together with the base VOC mixture in the presence of NOx. The concentration of formaldehyde is slightly underestimated by the model for reaction times after 13:00 GMT. However, this deviation reaches only a few ppb and is within the experimental error of the HCHO concentration (typically ± 20%). In total, the calculated and experimental concentration-time profiles are in good agreement. From Table 2 it is evident that the ozone formation rate of 82.7 ppb h-1 in the biodiesel experiment is the highest value of all runs. It should be pointed out that this is not specific for the combustion of biodiesel. The observation of Palm and Krüger (1998, 1999) that the combustion of biodiesel leads to higher formaldehyde emissions than the burning of a conventional diesel was not confirmed within the DIFUSO project (Wiesen 2000). The formaldehyde concentrations measured in the biodiesel exhaust were not higher than for all other diesel fuels tested. Accordingly, the high ozone formation observed in the biodiesel experiment from Nov. 18, 1999 was due to the experimental conditions. The photolysis intensity here was the highest of all the days of the campaign and the NO/NO2 ratio is the smallest of all the exhaust experiments performed. Both factors promote ozone formation in the system and are obviously the reason for the remarkably high ozone formation observed.

Investigation of Real Car Exhaust in Environmental Simulation Chambers

Figure 4.

291

Comparison of simulated () and experimental concentrations of NO (‹), NO2 (…) ozone (E) and formaldehyde (c) for the smog chamber run with fuel 2.

Sensitivity analysis with respect to ozone formation In order to identify the most important steps in the initial phase of the smog chamber runs leading to ozone formation and for further interpretation of the results, sensitivity analyses were carried out using the direct decoupled method of Dunker (1984). The sensitivities of ozone towards all rate coefficients of the chemical mechanism were calculated. For that purpose, rate parameter sensitivity coefficients, SK, were calculated as described by Stockwell et al. (1995). These rate parameter sensitivity coefficients depend on the concentrations of all reactants and, therefore, on reaction time. Thus time-dependent relative sensitivities can be calculated from the SK values. Figure 5 shows the relative sensitivities at t = 30 min (representing the initial phase of each experiment) of ozone towards selected reaction rate coefficients. These 9 reactions are the most sensitive for ozone formation and account for • 90% of the total sensitivity in the initial phase of each run. A positive sensitivity indicates that an increase of the rate coefficient will cause an increase of the ozone concentration, while increasing the rate constants with a negative sensitivity will lead to a smaller concentration of ozone. The plot illustrates the typical situation for a smog system: The photolysis of NO2 is the most important reaction leading to ozone formation, whereas the reaction of O3 with NO is the most significant sink for ozone in the system. Figure 5 clearly indicates that compared to the base mixture I experiment the formation of ozone in the exhaust studies is significantly promoted by the photolysis of the initial concentrations of nitrous acid and formaldehyde. For all diesel experiments, the relative sensitivities for these reactions are about a factor of 2 higher, whereas the importance of the NO2 photolysis and the reaction O3 + NO is lowered by a factor of about 2. The photolysis of higher aldehydes is of minor importance with respect to ozone formation.

P. Wiesen

292

It is remarkable that for all the reactions displayed in Figure 5, the corresponding sensitivities for the diesel exhaust runs are more or less equal, demonstrating that the different but significant higher ozone formation rates observed in comparison to the reference experiment are not specific for the diesel fuel formulation. They are clearly the result of more or less different initial parameters such as start concentrations or photolysis frequencies during the single experiments.

Figure 5.

Relative sensitivities of ozone towards the rate coefficients of selected elementary reactions of the chemical mechanism used at t = 30 min. The results for all diesel exhaust experiments are shown in comparison with those for the reference experiment with base mixture I. DCB = unsaturated dicarbonyls. For the calculation of the total fraction for the other reactions, the signs of the single sensitivities were not considered.

Summary and conclusions It has been demonstrated that atmospheric simulation chambers offer the unique possibility for studying the impact of real car exhaust under almost real-world conditions on tropospheric photosmog formation. The experiments performed in the EUPHORE smog chamber were simulated using a simple box model, including gas-phase chemistry only. The model described all experiments with high accuracy. Sensitivity analyses were carried out in order to explain the observations. The addition of diesel exhaust to a well-defined simple VOC mixture caused a significant increase of the ozone formation after irradiation, in comparison to smog experiments with a similar VOC/NOx ratio in the absence of exhaust gas. The increase of

Investigation of Real Car Exhaust in Environmental Simulation Chambers

293

ozone observed in the exhaust runs was mainly caused by the high initial concentrations of nitrous acid and formaldehyde emitted by the diesel engine. Higher aldehydes were of minor importance. The ozone formation was not dependent on the formulation of the diesel fuel. Differences in ozone formation rates for the single experiments were due to deviations in initial start concentrations as well as photolysis conditions. Further studies within the DIFUSO project studies exhibited that the influence of diesel soot on the ozone formation is negligible under the conditions of the smog chamber experiments performed. The present simulations using a simple gas-phase model leading to good agreement of experimental and calculated data confirmed this observation. For the future, detailed measurements of NOx, HONO and aldehydes in diesel exhaust for typical motor conditions should be considered. The use of such data as input parameters for the calculation of ozone formation potentials (e.g. incremental reactivities by the method of Carter et al. 1995) can contribute to a better estimation of the influence of diesel exhaust on tropospheric ozone formation. Acknowledgements Financial Support of this work by the European Commission, contracts N° ENV4CT95-0015 (INFORMATEX) and N° EV4V-CT97-0390 (DIFUSO), is gratefully acknowledged. The authors are deeply indebted to K. Wirtz, CEAM, Valencia (Spain), for kindly providing the experimental data, which were used for the model calculations performed in this study. References Atkinson, R.; Gas-phase tropospheric chemistry of organic compounds, J. Phys. Chem. Ref. Data, Monograph No. 2. (1994). Balek, S., R. Heck and S. Roth, S. (eds.).; Diesel engine combustion processes and emission control technologies, Society of Automotive Engineers, Inc., SAE Paper SP-1246, ISBN 1-56091-958-2, Warrendale (1997). Barnes, I. and J. Wenger (eds.); EUPHORE Report 1997, University of Wuppertal, Wuppertal (1998). Becker, K.H. (ed.); Influence of fuel formulation on atmospheric reactivity of exhaust gases (INFORMATEX), final report of the EC project, contract N° ENV4-CT95-0015, Wuppertal (1998) (see also: http://www.physchem.uni-wuppertal.de/PC-WWW_Site/Publications/Publications.html). Becker, K.H. (ed.); The European photoreactor EUPHORE, Final report of the EC project, contract N° EV5VCT92-0059, Wuppertal (1996) (see also: http://www.physchem.uni-wuppertal.de/PCWWW_Site/Publications/Publications.html). Becker, K.H.; The atmospheric oxidation of aromatic hydrocarbons and its impact on photooxidant chemistry, in: Borrell, P.M., P. Borrell, C. Cvitas, W. Seiler, Proc. EUROTRAC Symp. ’94, SPB Academic Publ., The Hague (1994) 67-74. Brown, P.N., G.D. Byrne andA.C. Hindmarsh; VODE: A variable-coefficient ODE solver, Journal of Science, Statistics and Computers 20 (1989) 1038-1051. Calvert, J., R. Atkinson, K.H. Becker, R.M. Kamens, J.H. Seinfeld, T.J. Wallington and G. Yarwood; Mechanisms of atmospheric oxidation of aromatic hydrocarbons, Oxford University Press, Oxford (2000). Dunker, A.; The decoupled direct method for calculating sensitivity coefficients in chemical kinetics, J. Chem. Phys. 81 (1984) 2385-2393. Gear, C.W.; Numerical initial value problems in ordinary differential equations, Prentice-Hall series in automatic computation, Vol. 17, Prentice-Hall, Englewood Cliffs (1971). Geiger, H., J. Kleffmann and P. Wiesen; Smog Chamber Studies on the Influence of Diesel Exhaust on Photosmog Formation, Atmos. Environ. 36 (2002) 1737 –1747. Geiger, H., K.H. Becker and P. Wiesen; Effect of Gasoline Formulation on the Formation of Photosmog: A Box Model Study, J. Air & Waste Manag. Assoc. 53 (2003) 425 –433.

294

P. Wiesen

Madronich, S.; Photodissociation in the atmosphere: 1. Actinic flux and the effects of ground reflections and clouds. J. Geophys. Res. 92 (1987) 9740-9752. Marshall, E.L. and K. Owen (eds.); Motor gasoline, Critical reviews on applied chemistry, Vol. 34, The Royal Society of Chemistry, Cambridge (1995). Palm, W.U. and H.U. Krüger; Comparison of ozone formation from diesel exhaust and rapeoil-methylester (RME): First results of smog chamber experiments, in: Borrell, P.M., P. Borrell, Proceedings of the EUROTRAC-2 Sympposium ’98, WIT Press, Southamptom (1999) 199-204. Palm, W.U. and H.U. Krüger (eds.); Experimentelle Untersuchungen des Ozonbildungspotentials von Motorabgasen bei Verwendung von Dieselkraftstoff und Rapsölmethylester, final report of the project "Ozonbildungspotential", sub-project 2, contract N° 95NR126-F, Fraunhofer Institute for Toxicology and Aerosol Research, Hannover (1998). Seefeld, S.; Laboratory kinetic and atmospheric modelling studies of the role of peroxyacyl nitrates in tropospheric photooxidant formation, Ph.D. thesis, Swiss Federal Institute of Technology, Zürich (1997) Seefeld, S. and W.R. Stockwell; First-order sensitivity analysis of models with time-dependent parameters: An application to PAN and ozone, Atmos. Environ. 33 (1999) 2941-2953. Stockwell, W.R., F. Kirchner, M. Kuhn and S. Seefeld; A new mechanism for regional atmospheric chemistry modelling, J. Geophys. Res. 102 (1997) 25847-25879. Stockwell, W.R., J.B. Milford, D. Gao and Y.J. Yang; The effect of acetyl peroxy-peroxy radical reactions on peroxyacetyl nitrate and ozone concentrations, Atmos. Environ. 29 (1995) 1591-1599. Wiesen, P. (ed.); Diesel fuel and soot: Fuel formulation and its atmospheric applications (DIFUSO), final report of the EC project, contract N° EV4V-CT97-0390, Wuppertal (2000) (see also: http://www.physchem.uni-wuppertal.de/PC-WWW_Site/Publications/Publications.html).

The EUROCHAMP Integrated Infrastructure Initiative Peter Wiesen Bergische Universität Wuppertal, Phys. Chemie / Fachbereich C, D-42097 Wuppertal, Germany Key Words: Atmospheric processes, Infrastructure, Environmental chambers

Introduction The ability of Europe’s research teams to remain at the forefront of all fields of science and technology depends on their being supported by state-of-the-art infrastructures. The term “research infrastructures” refers to facilities and resources that provide essential services to the research community in both academic and/or industrial domains. In this context the European Commission in Brussels is supporting research infrastructures in the framework 6 specific programme on structuring the European research area. The overall objective of this activity is to promote the development of a fabric of research infrastructures of the highest quality and performance in Europe, and their optimum use on a European scale based on the needs expressed by the research community. Specifically this will aim at: – ensuring that European researchers may have access to the infrastructures they require to conduct their research, irrespective of the location of the infrastructure; – providing support for a European approach for the development of new research infrastructures, also at the regional and transregional level, and for the operation and enhancement of existing infrastructures, including where appropriate facilities of worldwide relevance not existing in Europe. Five schemes for support are available under the Research Infrastructures action: 1. 2. 3. 4. 5.

Transnational Access, Integrating Activities, Communication Network Development, Design Studies, and Construction of New Infrastructures.

The objective of the Integrating Activities scheme is to support the integrated provision of infrastructure related services to the research community at a European level. It is also intended to have a structuring effect on the fabric of European research by promoting the coherent use and development of infrastructures in the fields it covers. To that end, the main characteristic of Integrating Activities will be their capacity to mobilise a large number of stakeholders in a given class of infrastructure. The ambition of such Integrating Activity should be to induce a long-term integrating effect on the way research infrastructures operate, evolve and interact with similar infrastructures and with their users, thereby contributing to structure the European Research Area: x Operators of similar infrastructures in a given class should for instance find it easier, through an Integrating Activity, to develop synergies and complementary capabilities in such a way as to offer an improved access to researchers. 295 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 295–299. © 2006 Springer. Printed in the Netherlands.

296

P. Wiesen

x Likewise, infrastructure operators and users should be in a better position to tackle new or unexpected developments in their field, for instance in relation to state-of-the-art instrumentation, with a more co-ordinated approach. x More generally, a closer interaction between a large number of scientists active in and around a number of infrastructures will facilitate cross-disciplinary fertilisations and a wider sharing of knowledge and related technologies across fields and between academia and industry. This scheme is to be implemented primarily through so-called Integrated Infrastructure Initiatives such as the new EUROCHAMP project, which is presented here. The project The EUROCHAMP project integrates the most important environmental reaction chambers in Europe for studying atmospheric processes into a Europe-wide infrastructure. These facilities were created by multinational initiatives to study the impact of atmospheric processes on regional photochemistry, global change, as well as cultural heritage and human health effects under most realistic conditions. Although initial advances in the application of large chambers occurred in the United States, Europe now leads the world in the use of large, highly instrumented chambers for atmospheric model development and evaluation. Smaller chambers that were designed for specific purposes and are operated by experts in their fields excellently support such chambers. The integration of all these environmental chamber facilities within the framework of the EUROCHAMP infrastructure promotes retention of Europe's international position of excellence in this area and is unique in its kind worldwide. The mobilisation of a large number of stakeholders dealing with environmental chamber techniques provides an infrastructure to the research community at a European level that offers a maximum support for a broad community of researchers from different disciplines. The EUROCHAMP project initiates a currently not existing structuring effect of atmospheric chemistry activities performed in European environmental chambers, since it offers the full availability of corresponding facilities for the whole European scientific community. The consortium of partners, which is described below, provides their expertise and experience in atmospheric chemistry to researchers of different disciplines, policy and industry and offers an infrastructure that can be used by interested parties for solving a large variety of problems related to atmospheric science. Grid of partners and associated user groups Besides the 12 partner institutions, a number of selected associated user groups with a high grade of expertise in the different fields of interest provide their experience either as advisers to special topics or as potential users of the infrastructure. The partners and associates institutions are widely distributed over Europe. Two associated user groups are located in the United States of America. Figure 1 shows a map in which the locations of each participant are shown.

The EUROCHAMP Integrated Infrastructure Initiative

Figure 1.

297

Institute locations of the partners and associated user groups of the EUROCHAMP infrastructure. For explanation of the abbreviations please refer to http://www.eurochamp.org.

EUROCHAMP objectives and activities Three networking activities and two joint research activities are formulated and crosslinked in the EUROCHAMP project. Networking activities The major objective of the networking activities within the EUROCAHMP project is the initiation of an effective interdisciplinary collaboration between the community of

298

P. Wiesen

atmospheric scientists and colleagues from other disciplines that are closely related to it. This will be achieved through the three networking activities of EUROCHAMP. Networking activity N1 The objective of networking activity N1 is the generation and application of standardised rules as a method of quality assurance for raw data analyses of the experiments in each facility. For this purpose a number of inter-comparison studies applying analytical devices in reference experiments will be carried in the different chambers, which provides an indirect measure of the infrastructures' excellence. Networking activity N2 In order to make the results of experiments performed in the partners’ facilities most transparent and accessible to the scientific community, a standardised data protocol for chamber studies will be defined. This standardised form will be the basis for the central database of environmental chamber studies to be constructed within the project. This WWWbased database will be made accessible to the whole scientific community, leading to a very effective dissemination of the results. Networking activity N3 Within networking activity N3 four larger international conferences/workshops on infrastructure-related topics will be organised. In order to reach a maximum of success, internationally established experts on the corresponding topics will be invited to join these conferences. The results will be published in suitable proceedings for dissemination to the scientific community. Joint research activities The major objective of the joint research activities within the EUROCHAMP project is the optimisation and further development of the infrastructures' performance. In order to meet these goals, two corresponding research activities are defined in the EUROCHAMP work programme, namely the development and refinement of analytical equipment and the development of chemical modelling techniques. Joint research activity JRA1 The development of novel and the refinement of existing analytical devices of environmental chambers in order to successfully detect atmospheric trace species or to characterise aerosol particles is an essential task to be followed over the whole lifetime of the EUROCHAMP research facilities. The increasing demands for more comprehensive analytical techniques caused by the more and more complex scientific questions to be answered, requires a continuous improvement of the technical possibilities of a chamber. Accordingly, the project includes a number of research activities focused on this topic: x characterisation of oxygenated volatile organic compounds (OVOCs), x radical measurements (OH, HO2, RO2), x nitric acid measurements, x characterisation of aerosols. Besides the optimisation of existing devices, a number of analytical devices will be completely newly designed and introduced for the first time in an environmental chamber. The highly specific equipment will be developed in a mobile form, so that such instruments may be transported to a chamber of choice and used in selected experiments independent of

The EUROCHAMP Integrated Infrastructure Initiative

299

localisation. This philosophy strengthens the idea of a real grid of environmental chambers forming a powerful infrastructure. In addition, the instruments to be developed will be of great use for future field campaigns for which sophisticated, improved analytical instrumentation is urgently required. Joint research activity JRA2 The field of chemical modelling is directly coupled to each type of environmental chamber studies. The analysis of chamber experiments without any model application is mostly not possible. Accordingly, model activities are urgently necessary and a permanent companion of each experimental task. 1. Since the quality of simulation studies strongly depends on the question how exactly the chemical behaviour of the chamber itself is characterised, sensitive parameters urgently st required for simulations have to be determined for each facility of the infrastructure (1 objective). The dissemination of each result from such studies serves for a better interpretability of environmental chamber studies as a whole. 2.

The second objective for model applications is the test of complex chemical mechanisms used for multi-phase model applications related to chamber experiments. An established chemical code for interpretation of chamber studies that can be applied to all facilities increases the quality of the whole infrastructure and offers new possibilities for solving open questions of interest for the European researchers’ community.

3.

Furthermore, chemical models to be developed can be applied in the EUROCHAMP network for the solution of specific problems in atmospheric chemistry, e.g. development and validation of degradation mechanisms of organic pollutants that are of paramount importance, the investigation of atmospheric reactivity as an overall property under various conditions or the influence of alternative fuels or solvents as well as bio rd fuels on tropospheric chemistry (3 objective).

In conclusion, the EUROCHAMP project strengthens the idea of a real grid of environmental chambers as powerful tool for system analysis increasing the value of the whole chambers infrastructure for the European research community. Detailed information, e.g. description of the partner institution, the available environmental chambers and their analytical equipment, events such as meetings and workshops etc. can be found under http://www.eurochamp.org.

Survey on Atmospheric Chemistry Research in Some New EU Member States and Candidate Countries Ekaterina Batchvarova, Tatiana Spassova, Nedialko Valkov, and Liliana Iordanova National Institute of Meteorology and Hydrology, Blvd Tzarigradsko Chaussee No. 66, 1784 Sofia, Bulgaria [email protected] Key Words: Atmospheric chemistry, Air quality monitoring, Air pollution modelling, Atmospheric aerosol, Ozone

Introduction Historically, some of the new EU Member States and the Candidate Countries have experienced high levels of pollution in the past. Enhanced management measures were and are needed to improve their air quality. A survey was recently conducted (Batchvarova et al., 2005) to list the current research activities on atmospheric chemistry in these counties, as well as the groups and institutions involved in it. Air chemistry is an essential element of air quality, climate change modelling, industrial energy planning, and health risk assessments. Atmospheric chemistry research is also related to air quality monitoring. Therefore, the survey has also retrieved information on air quality monitoring networks and their management of those countries involved. Some information on air pollution modelling research is also discussed. The ongoing research (field, laboratory and modelling) in the field of chemical transformation of trace compounds in the atmosphere is discussed here and parallels are drawn among 10 of the new EU Member States and Candidate Countries, namely Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia. Laboratory studies traditionally lay emphasis on rate and equilibrium processes. Field studies are based on aircraft and surface measurements of reaction chemistry, advective influences on the chemical composition of the atmosphere, and air-surface exchange processes. Both field and laboratory experimental studies on atmospheric chemistry are very demanding on equipment and resources. Therefore, most of the funding for these studies is coming from international projects (EC, ESF, NATO, etc.). Modelling efforts address both chemistry and dynamics on regional and global scales. The analysis of research activities in these fields is made with regard to current EU (European Union) practice and the historical frameworks in the ten countries of interest. The unique traditions and achievements in atmospheric chemistry research in these countries are highlighted. The participation of research teams in different projects and initiatives over the last 5-10 years of integration of these countries within the European Framework Programmes is surveyed. Since the 1990s, a number of surveys on environmental quality and management were prepared as part of different PHARE missions and tasks, as part of the European Environmental Agency (EEA) activities and as pre-accession twinning projects under the Twinning Programme of the EC (European Commission). Yet, the research activities within the 10 countries were not discussed. Through the FP4, FP5 and FP6 (Framework Programmes) of the EC the research communities of these countries had the possibility to participate in European-wide scientific projects. There were collaborative achievements, but much less than expected and desired as pointed out in analyzes of the FP4 and FP5 results. The investigations on air pollution modelling

301 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 301–340. © 2006 Springer. Printed in the Netherlands.

302

E. Batchvarova et al.

and the atmospheric chemistry are closely connected to environmental management, but did not get special attention from national funding. The present survey places emphasises on the atmospheric chemistry research in the 10 countries named above. It also describes the air quality monitoring systems and air pollution modelling studies in these countries. The aim is to gather information and make it available in condensed form in order to facilitate collaboration with other nations and also strengthen the research within the countries. Research on atmospheric chemistry is not an essential part of the overall chemical research in almost all of the 10 countries. Only a few groups from the Czech Republic, Poland, Estonia, Slovenia, Latvia, Lithuania and Hungary are dealing with aerosol studies and chemical processes in the atmosphere. The rest of the chemical research is in other fields and only partly concerns the atmosphere. The survey is structured in detailed discussions arranged by country and topics. Most abbreviations are introduced and explained in the text. The parallels, analysis and synthesis of the information are given separately. Air quality monitoring was started in all 10 countries (Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia) in the 1950s. Initially, the radioactivity in the air and dry and wet fallout were considered. The task was given to Atomic Agencies, Institutes of Hygiene and Meteorological services to organize monitoring networks. In the 1970s chemical air pollution monitoring was started as well. Typically, the Hygiene Institutions belong to Ministries of Health and the Meteorological services are attached to different structures. For decades the monitoring networks were developed separately in the different institutions with some differences in observation programmes that reflect the different scopes of their activities and objectives. The meteorological services have a climatological and synoptic monitoring approach, while the hygiene institutes are interested in non-regular sampling for air quality in residential and industrial areas. During the late 1970s state agencies (committees) were established to deal with the environmental protection by controlling industrial emissions. Gradually, in some countries these agencies also developed air quality monitoring networks and were authorized to gather all the information and issue Bulletins (monthly, quarterly or yearly). The atmospheric chemistry research, such as development of sampling and analysis methods, was concentrated in the air pollution monitoring laboratories. Chemical research in a broader sense was concentrated in the chemical institutes in the Academies of Sciences and in universities. Typically, very limited communication and collaboration among the different institutes was established. The reason was probably that the financing system of research was separate for the different institutions. Most of the common activities were due to personal contacts between scientists. The economical and political changes in all these countries since 1989-1990 have lead to transformation of their air quality monitoring networks in different ways. Finally, as all countries join the European Union, the result in terms of air quality monitoring principles is and will be very similar. The organization of science has also changed differently. In Bulgaria, Romania, Poland, Hungary and the Czech Republic the Academies of Sciences were retained but reorganized. In Estonia, Latvia, Lithuania, Slovenia and the Slovak Republic, the situation was different due to their former associations with a larger system and also to the size of the countries and population.

Survey on Atmospheric Chemistry Research in Some New EU Member States

303

The parallels that can be drawn among the countries are to be found in the historical development of all the 10 countries during the last 50 years. Future requirements will be harmonized being based on the European Union environmental legislation and also on the wide international collaborations necessary in the field. Overview of Air Pollution Monitoring, Modelling and Atmospheric Chemistry Research by Country The surveys by country include an overview of Air quality monitoring, Atmospheric Chemistry research and Air pollution modelling. For all topics, Institutions and groups involved, Historical aspects, Current projects and studies performed in the last 5 years, National strategies, Achievements, and Publications are discussed. An analysis of problems and the identification of needs is made collectively for all the countries. Bulgaria Ambient Air Quality Monitoring Network Air quality monitoring in Bulgaria started in 1975 (Parvanova, 2002) and was performed by the Institute of Hygiene at the Ministry of Health, the National Institute of Meteorology and Hydrology (NIMH) at the Bulgarian Academy of Sciences and the newly established, at that time, Committee on Environmental issues at the Government. The ambient air radioactivity measurements started at NIMH in the 1950s. In the 1970s significant transformations and changes were made in the institutional structure of the monitoring system. Along with the need for emissions’ control a governmental body was needed information wise, although the environmental information was not publicly available. After 1995 the network was supplied with sophisticated sampling and laboratory equipment through the EC PHARE programme, all concentrated at the Ministry of Environment and Waters (MEW). The air quality monitoring and the emission inventory are financially guaranteed mainly by the state budget and partially by the National Fund at MEW. Because of economic difficulties, the allocated finances from the budget for air quality monitoring are limited, which has resulted in a decrease in the number of manual stations since 1995. The lack of funding is more severe at institutions outside the MEW, like NIMH. A database information system was started in January 1986. Monthly averaged data had been received from the institutions, responsible for the monitoring. At NIMH the data are available in electronic format since 1982 and they are a part of the WMO Green House Gases Database in Tokyo, Japan. Presently, the air quality data management is the responsibility of the EAE (Executive Agency on Environment). There, a new database was created in 1991. Input data are collected from the Regional Environmental Inspectorates (REIs), EAE, NIMH and the Hygienic Inspections. The network consists of 68 stationary stations - 16 automated (on-line) and 52 with manual sampling and chemical analysis, as well as 6 mobile automated stations. The stations are located in 37 settlements spread over the country - urban, residential, high traffic and industrial areas. All the manual stations operate in a unified sampling regime and standardized analytical procedures in accordance with Bulgarian National Standard “Basic rules for air quality sampling”. The sampling frequency is 4 times per day, 5 days per week. The basic measured parameters are: TSP, Pb, aerosols, SO2, NO2, and H2S. In relation with specific industrial

304

E. Batchvarova et al.

activities, additional pollutants are also measured, such as: NH3, phenol, arsenic aerosols, HCl, Cl2, HF and fluorides. There is one National Background Station that is part of the GEMS system of UNEP, WMO and UNESCO. It is located on Rojen mount, one of the highest mountain peaks in Bulgaria. The station works on a limited measurement programme due to a lack of governmental funding to equip it. Six OPSIS systems operating continuously are situated in the most affected industrial areas in the country. In addition another seven OPSIS systems are working in the Danube area, mainly in relation with trans-boundary air pollution between Bulgaria and Romania. The establishment of a network along the river Danube was partly financed by a PHARE project. In the air quality laboratories of EAE and REI chemical analyses are performed in accordance with standardized analytical methods. A Quality Handbook, which is periodically updated, is also available. Control of the measurement accuracy is performed (calibration curve settings, verification with standard samples prepared in the laboratory). Comparative assessment of the results of the chemical analyses from different measuring devices is performed. The laboratories of EAE are accredited by the Bulgarian Accreditation Body. Air quality data management is another responsibility of EAE. Raw and aggregated data are stored in local databases of all REIs. The data from REIs are sent to the national database at EAE. NIMH participates in the air quality network with 5 stations, where 24-hour data are gathered in connection with synoptic measurements and thus relevant 24-hour means are delivered to the WMO Data Base for greenhouse and reactive gases. The Air Quality Framework Directive (96/62/EC) provides a comprehensive strategy for the management of air quality in EC Member States linking controls on emissions with the attachment of the air quality objectives. This Directive is progressively taking effect as its daughter Directives are adopted and enter into force. It is transposed into Bulgarian legislation mainly through Regulation N7 on ambient air quality assessment and management. The Bulgarian air quality standard legislation is drafted with the aim of transposing forthcoming EU legislation. Regulation N9 on limit values for sulphur dioxide, nitrogen dioxide, particulate matter and lead in ambient air should fully transpose the first daughter Directive to the Air Quality Framework Directive. The air quality standards are well covered by Bulgarian Legislation, and no additional instruments are needed. It can be noted that some of the air quality standards before 1990 were stricter than those of the EC. Directive 92/72/EEC on tropospheric ozone pollution is also fully transposed into Bulgarian legislation. The main instrument is Regulation N8 on ambient air quality limit values for ozone. The Bulgarian requirements for the ozone-monitoring network comply well with the provisions of the Directive. Atmospheric Chemistry Research At the University of Chemical Technology and Metallurgy (UCTM), the Faculty of Ecology and environmental protection, among the issues for education and research are - the industrial waste utilization and natural resources and their protection monitoring. There is a close co-operation between the UCTM’s Scientific Research Centre and all Bulgarian scientific

Survey on Atmospheric Chemistry Research in Some New EU Member States

305

institutions, as well as with related Universities in Europe, Asia and America. UCTM takes an active part in nearly all European scientific programmes – ERASMUS, COPERNICUS, COST, 5th and 6th Framework Programmes, etc. The University of Plovdiv “Paisii Hilendarski”, Faculty of Chemistry is dealing with development and application of new analytical methods for environmental research. At the Sofia University “Sv. Kliment Ohridski”, Faculty of Chemistry, Department of Analytical Chemistry the group of Prof. Vasil D. Simeonov is performing research in analytical chemistry, chemometrics, environmetrics, multivariate calibration; classification, interpretation and modelling of environmental data sets; evaluation and optimization of analytical procedures; potentiometry with ion selective electrodes; atmospheric and marine chemistry. A very wide network of international collaboration is associated with the group. At the Bulgarian Academy of Sciences (BAS) there are a number of research institutes on chemistry. Among them the Institute of Physical Chemistry, the Institute of Catalysis, and the Institute of General and Inorganic Chemistry work on topics close to atmospheric chemistry, but no specific projects were recently undertaken in this field. Air Pollution Modelling and Other Topics Experienced teams of air pollution modellers are working at the National Institute of Meteorology and Hydrology (NIMH) and at the Geophysical Institute (GPhI). Both institutes belong to the Bulgarian Academy of Science. Several advanced national air pollution models, including 3D Eulerian local-to-regional model, are available. Additionally, the pre-processing possibilities are relatively good, e.g. a 3D hydrostatic prognostic model was implemented and a tradition in boundary-layer meteorology has been established. The most significant current project is BULAIR, an EU FP5 Project EVK2-CT-200280024 (BULAIR, 2002). Institutions involved are NIMH and GPhI. Specialists from other BAS institutes as well as experts from the Ministry of Environment and Waters, National Statistical Institute, etc. are involved as well. The scope and objectives of the project are the creation of a database - collection, evaluation and organization of data; Comprehensive review of existing modelling tools, choice of the most relevant for use in the project; Staff training; Increased networking and twinning activities with EU centres. The BULAIR project is the tool to bring the air pollution modelling studies, which started in the 1970s, to an up-to-date technology level. Also the concentration of efforts and collaboration of institutions within Bulgaria and in Europe are an important part of BULAIR. NIMH and GPhI participated in a number of other European projects starting with the EC FP4. MEDCAPHOT-TRACE (EV5V-CT93-0312) was a Mediterranean campaign of photochemical tracers – transport and chemical evolution, which was a comprehensive field experiment study in Athens in 1994. ETEX (The European Tracer Experiment) and RTMOD (the Real-Time Modelling) gathered the modelling efforts of a number of European groups. Scientists from NIMH are participating in the FP5 ENSEMBLE and MERLIN and the FP6 ACCENT and QANTIFY projects. A number of NATO Science for Peace scientific and linkage grants as well as extensive participation in COST 715 were also part of the activities. In urban boundary-layer meteorology a Swiss-Bulgarian bilateral Institutional partnership took place in 2001-2003.

306

E. Batchvarova et al.

Czech Republic Ambient Air Quality Monitoring Network The main services that carry out air quality monitoring are the National Institute of Public Health (NIPH) and the Czech Hydrometeorological Institute (CHMI). In addition a number of companies and institutions contribute to the national network with a few stations each. The National Institute of Public Health, Department of Air Hygiene supports an Expert Group for Ambient Air Hygiene. The activities comprise the processing of expert opinions, health risk assessments, legislative activities, tuition, and consulting. In 2003, urban air pollution was monitored at 76 stations (44 and 32 operated by the Ministry of Health and the Ministry of the Environment, respectively) located in 27 cities involved in the Monitoring System (SZU, 2004). In 2003, sulphur dioxide (SO2), nitrogen oxides (NO/NO2/NOx), particulate matter (TSP and/or suspended PM10 fractions), and mass concentrations of selected metals (arsenic, chromium, cadmium, manganese, nickel and lead) in particulate matter samples were monitored in all the cities of the Monitoring System except for MČlník. The SO2 measurements in the Public Health Service network were terminated at all the manual stations in the cities with CHMI stations; in the cities without a CHMI station, measurements are made during the heating season only). Concentrations of carbon oxide, ozone, polycyclic aromatic hydrocarbons (PAHs) and volatile organic compounds (VOCs) continue to be monitored selectively in a number of the monitored cities. The Czech Hydrometeorological Institute (CHMI) takes care of the national ambient air monitoring system. The charter of CHMI was amended in June 1994 and in August 1995 pursuant to a decision adopted by the Minister of the Environment of the Czech Republic (MoE CR). The amendment defines the aim, object, and functions of the Institute. The Czech Hydrometeorological Institute is a central State institute of the Czech Republic in the fields of air quality, hydrology, water quality, climatology, and meteorology. The objectives of CHMI's activities include: i) the integration of the performance of public service in a rational, efficient and economical manner, ii) the establishment and operation of monitoring stations with the aid of a telecommunications network (national networks for monitoring the atmosphere and hydrosphere both qualitatively and quantitatively and for establishing the reasons underlying their pollution or damage), iii) processing in an expert manner the results of observations, measurements and monitoring, iv) the creation and maintenance of data bases, and v) the provision of information on conditions, characteristics and regimes. The monitoring system of the CHMI consists of 97 automatic stations. The measured pollutants are SO2, NO, NO2, PM10. In some stations O3, CO, organic species such as benzene and toluene are measured. Measurements by airplane L410 Turbulent have been performed since 1979. In the last 10 years the programme has been focused mainly on tropospheric ozone. The measurements include SO2, NO, NO2, NOx, O3 concentrations and temperature at levels from 150 m above ground to 3500 m above sea level.

Survey on Atmospheric Chemistry Research in Some New EU Member States

307

Detailed graphical and other information about the Czech air pollution-monitoring network is provided by CHMI, 2004. Two stations, Svratouch and Kosetice, are the Czech contribution to the EMEP monitoring network. They work on a full programme of measurements for precipitation (sulphate, nitrate, ammonium, magnesium, sodium, chloride, calcium, potassium, conductivity, pH), air (sulphur dioxide, nitrogen dioxide, nitric acid, ammonia, ozone, sulphate, nitrate, ammonium, sum of nitric acid and nitrate, sum of ammonia and ammonium). Atmospheric Chemistry Research The Institute of Chemical Process Fundamentals, Academy of Sciences of the Czech Republic hosts the leading Aerosol Laboratory in the country. The main effort is aimed at experimental study of individual parts of the condensation process (nucleation, condensation and evaporation, heat and mass transfer) as well as at more complex phenomena such as gas-phase synthesis of nano-particles, and combustion or atmospheric aerosols. The Institute has experience regarding the development of measurement methodology for air pollution: the pre-concentration of pollutants, measurements of NH3, SO2, NO2, HNO2, HNO3 and O3 determination in ambient air. The group has extensive long-term experience in field measurements and in pollutant monitoring campaigns. The department owns: unique analyzers for monitoring ozone, nitrogen dioxide and nitrous and nitric acids. The fields of research are: particulate emissions from combustion processes; composition and size of atmospheric aerosols; indoor/outdoor aerosols; nucleation phenomena; synthesis of nano-particles via aerosol processes; heat and mass transfer in aerosol systems; and interaction of aerosols with electromagnetic radiation. Some research projects are supported by the EC (“Sub grid scale investigations of factors determining the occurrence of ozone and fine particles” - grant No. EVK2-CT-199900052 SUB-AERO; “Characterization of urban air quality – indoor/outdoor particulate matter chemical characteristics and source-to-inhaled dose relationships” - grant No. EVK4-CT-00018 URBAN-AEROSOL; “Integrated exposure management tool characterizing air pollution-relevant human exposure in urban environment” - grant No. EVK4-CT-2002-00090 URBANEXPOSURE). Other studies are funded by the National Grant Agency or Ministry of Education (“Composite nano-particle synthesis by the CVD method in a hot-wall tube flow reactor” - grant No. 104/02/1079; “Aerosol particle growth in presence of foreign gas and problem of foreign molecule trapping” - No. IAA4072205; “Development of experimental methods for measurement of nucleation rates in mixtures present in clean and polluted atmospheres” - grant No. 2076203; “Comprehensive size resolved characterization of atmospheric particulate matter in Prague” grant No. 205/03/1560; “Nucleation studies using diffusion chambers. Atmospheric aerosol measurements - instruments inter comparison” - grant No. ME699, program KONTAKT; “Indoor aerosol deposition: an experimental study” - grant No. 101/04/1190, 2004-2006 [6318]; “Composition and mode of occurrence of the mineral constituents in brown coal and their behaviour during fluidised bed combustion”; “Reduction of heavy metal emissions from fluidised bed coal combustion using sorbents”; “Physicochemical properties of urban atmospheric aerosol. Source apportionment and impact on air quality”). The research of the groups at the J. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Department of Molecular Simulations of

308

E. Batchvarova et al.

Atmospheric Aerosols, is in atmospheric reactivity of sea-salt aerosols and urban air pollution control. Measurements of pollutants, physical and chemical modelling of secondary pollutants production and propagation in the urban and rural areas are performed. The main research activities of the J. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Department of Chemical Physics, partly concerning atmospheric chemistry are within Gas Phase Ion Chemistry (experimental and theoretical studies of the dynamics and kinetics of processes involving singly- and multiplycharged ions; dynamics of ion-surface collisions; studies of ion-molecule reaction kinetics relevant to medicine-related microanalysis; application of mass spectrometry to structural studies of organic, bioorganic and organo-metallic compounds, and of mechanisms of ionization and fragmentation processes). Research of fog and rime water chemistry and physics is performed at the Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic. The Czech Aerosol Society (CSA) was formed in 1999 from the former Working Group on Aerosol Research of the Czech Society of Chemical Engineering at the occasion of the European Aerosol Conference held in Prague in 1999. As given in its constitution the Society maintains a forum of researchers from various Czech institutions and universities in order to promote collaboration in all areas of aerosol research, to promote by means of meetings and publications the spread of information between the members and the public, to support education in aerosol related fields at all levels, and to support international co-operation. At the Research Centre of Environmental Chemistry and Ecotoxicology (RECETOX) - Air Sampling Group (AIRSAG), research is performed by staff of the centre as well as co-workers from TOCOEN Ltd. and M.Sc. and Ph.D students. Some recent projects: x EcoRA - Ecological Risk Assessment; x UNIDO Project - Implementation of Stockholm POPs convention in the Czech Republic Project objective is preparation of bases for ratification of Stockholm Convention; x National Inventory of POPs in the Czech Republic; x GEF UNEP POPs Project - Regional Report from Region III – Europe; x ELICC – Dioxins & PCBs: Environmental Levels and Human Exposure in Candidate Countries European Commission, Brussels REFERENCE: ENV.C.2/SER/2002/0085 x Transformation of Air Polluting Compounds (Project No. VaV/740/2/01, In Czech); x PBTs Project OCOEN REPORT No.150: Persistent, Bioaccumulative and Toxic Chemicals in Central and Eastern European Countries (State-of-the-art Report prepared under R-T&A by Ivan Holoubek, Anton Koþan, Irena Holoubková, JiĜí Kohoutek, Jerzy Falandysz); x Project Košetice TOCOEN REPORT No.194: Middle European Monitoring of PBT Compounds in Košetice Observatory, South Bohemia by Ivan Holoubek et al. The Project objective is long-term regular monitoring of selected POPs in air and in other environmental compartments in Košetice Observatory of CHMI.

Survey on Atmospheric Chemistry Research in Some New EU Member States

309

Air pollution modelling and other topics The most severe urban air pollution problem in the Czech Republic is the traffic air pollution in large cities, particularly in Prague. The Gaussian model developed at the Czech Hydrometeorological Institute (CHMI) is frequently used in urban air pollution studies. A Lagrangian photochemical model SMOG, with simple chemistry, was developed at the Charles University of Prague. An advanced 3D street canyon research model is being developed at the Technical University of Brno. The French modelling system Fluidyn Panache is tested at the Technical University of Ostrava. The US NOAA model HYSPLIT was recently implemented at CHMI. Prague participates in the international project HEAVEN, the main objective being the development of a system for traffic air pollution management in several large European cities. The Swedish modelling system AIRVIRO is to be used in the project. Recent implementation of the French limited area meteorological prognostic model ALADIN - LACE substantially increased pre-processing possibilities in the Czech Republic (e.g. 8 grid points in the territory of Prague). A relatively high quality emission database is available for the Czech Republic, particularly for Prague. Theoretical work and wind tunnel experimental work on boundary-layer theory is carried out at the Institute of Thermodynamics of the ASCR. Estonia Ambient Air Quality Monitoring Network At present, all the monitoring sites in Estonia are operated by the National Monitoring Centre at the Environmental Research Centre. Here the maintenance of the stations as well as a central calibration of the instruments is carried out. It is planned that regional authorities, like the town government of Tallinn, will have on-line access to the data of the monitoring sites of particular interest to them. According to the Final Report of a twinning Estonian-German project (Twinning, 2000) on assessment of the Estonian air quality monitoring system some additional monitoring sites were suggested mainly in the southern rural areas of the country. These stations could collect information on trans-boundary air pollution imported from the neighbouring countries. The Estonian air quality monitoring network would then consist of 14 monitoring sites. The information is published in the Estonian National Environmental Strategy (NES), Atmospheric Air, 2000. The Estonian National Monitoring Centre (NMC) is installed within the Estonian Environmental Research Centre (ERC) in Tallinn. All measuring results from monitoring sites are collected and processed at NMC using a software product of an Austrian company, as a part of a PHARE investment project. On this basis, experts of NMC prepare monthly and yearly reports on air quality and reports concerning certain evaluations of a specific interest. On-line presentation of information to the public is provided via the INTERNET. Within Estonia all monitoring of air pollution is carried out at the Estonian Environmental Research Centre. The Environment Information Centre (EIC) has two main tasks. It collects, processes, issues and stores environmental data. EIC transfers data and information to several Baltic, Nordic and European partners. It acts as the counterpart to the European Environmental Agency (EEA)

310

E. Batchvarova et al.

and reports, e.g., the emission data of the stationary sources. EIC prepares information and monitoring related acts, such as: The Environmental Register Act; Legal act on applying the Århus Convention; Regulation by the Minister of Environment on the Procedure of Implementing state environmental monitoring sub-programs. Formerly, EIC co-ordinated the state environmental monitoring programme, make the main contracts with different monitoring bodies and managed and processed collected data. Since September 1999, the University of Tartu was appointed to co-ordinate the National Monitoring Programme (with 16 sub programmes). This includes analyses of statistical data, production of trends and publication of data on the Internet. Atmospheric Chemistry research At the National Institute of Chemical Physics and Biophysics (NICPB), Laboratory of Chemical Physics the research activities range from studying fundamental aspects to applications. Strong emphasis is given to the deployment of modern physical methods in chemistry and biochemistry, which include: principles of mass-spectroscopy and ion-cyclotron resonance; environmental analyses; and theoretical and experimental particle physics. The Centre of Excellence of Analytical Spectrometry was formed in 2001 on the basis of six research groups of the Laboratory of chemical physics of NICPB and one biology-related research group. The laboratory also provides services as a reference laboratory, performs environmental monitoring and general chemical analyses. In addition, the facilities are used for advanced degree studies. Among several groups within the centre, those on Mass-Spectrometry and Catalysis and Environmental Chemistry work in the field of atmospheric chemistry. Experimental studies of oil shale combustion fly ash aerosol under simulated day-time and night-time atmospheric-like conditions were performed at NICPB in a 190m3 outdoor Teflon film chamber. Continuous monitoring of particle size distributions is performed with various optical and electrical devices. The respiration fraction of particles, which contributes most to the health effects of the aerosols, is studied quantitatively. NICPB works in close collaboration with institutions in Estonia, such as the University of Tartu, Institute of Environmental Physics; The Tallinn Technical University; The Estonian Environmental Research Centre and the Geological Survey of Estonia. Among the international partners are the European Science Foundation program INTROP (Interdisciplinary Tropospheric Research: from the Laboratory to Global Change, 2004 – 2008), German, Norwegian and USA institutions. An Electrical Aerosol Spectrometer (EAS) was developed in the Institute of Environmental Physics of the University of Tartu as a result of the 20-year research and developing work. The latest version of EAS has been built by AIREL Ltd. Tartu University, Institute of Technology is a multidisciplinary research institution. Several groups are working on atmospheric pollution topics, such as The Aerosol Research Group; the Atmosphere Dynamics Work Group; the Environment Technology Work Group; the Environmental and Occupational Health Work Group and the Environmental Radioactivity Work Group.

Survey on Atmospheric Chemistry Research in Some New EU Member States

311

Air pollution modelling and other topics Decreasing SO2 concentrations and increasing NOx and CO concentrations are observed in Estonian towns. Such trends, which are typical for many East European towns, result from the increasing traffic and from the increasing consumption of natural gas. The urban air pollution model AEROPOL (based on the Gaussian concept) was developed at the University of Tartu, Institute of Environmental Physics of the University of Tartu. The model has been used to estimate urban air pollution in several Estonian towns. The AEROPOL model was tested against a Norwegian (Lillestrom) data set. Recently, the Swedish model system AIRVIRO was implemented in Tallinn. Parameterization of dry deposition processes is among the most recent study activities. Hungary Ambient Air Quality Monitoring Network The organizations, which are currently involved in air quality monitoring, are the Ministry of Health and Social Welfare and the National Meteorological Service. The Ministry for Health and Social Welfare has the responsibility for measuring ambient air quality throughout Hungary. In order to accomplish this task, nitrous oxides, sulphur oxides and particle sedimentation are regularly monitored at 100 stations. The National Meteorological Service maintains the background monitoring network consisting of 4 stations (K-pusta, Farkasfa, Nyírjes and Hortobágy) in accordance with the recommendations and quality requirements of the EMEP. Measurements of the gases (sulphur dioxide, nitrogen dioxide, ozone, ammonia, nitric acid), aerosol particles (sulphate, nitrate, ammonium, sodium, potassium, calcium) and precipitation (pH, conductivity, sulphate, nitrate, chloride, ammonium, sodium, potassium, calcium) are performed. In addition, methodological guidance is provided for the operation of two further stations run by the local Environmental Protectorates: Fertoujlak and Majlatpuszta. Wet deposition of nitrogen and phosphorus over Lake Balaton is estimated by means of precipitation samples collected at Siofok and Keszthely. Sampling and laboratory analyses are performed on the basis of EMEP and WMO GAW Technical Manuals. Besides the operative activities, experts are involved in several national and international projects such as: evaluation and replacement of National Standards for atmospheric transmission; development of accidental (nuclear) dispersion models; deposition of nitrogen and phosphorus over Lake Balaton; three EUREKA/EUROTRAC-2 Projects (BIATEX-2, SATURN, TOR-2) and COST 615 and 715. Atmospheric Chemistry research At the Eötvös Loránd University - Department of Chemical Technology and Environmental Chemistry, a coupled Eulerian photochemical reaction–transport model and a detailed ozone dry-deposition model for the investigation of ozone fluxes over Hungary have been developed and are used in collaboration with Leeds University of the UK. As part of a research project with Ghent University, aerosol samples were collected using several filter-based devices (Nuclepore polycarbonate membrane, Teflon membrane and quartz fibre filters) over

312

E. Batchvarova et al.

separate daylight periods and nights, and on-line aerosol measurements were performed by TEOM and aethalometer. The study was carried out in an urban canyon and also at a near-city background site in Budapest, Hungary from 23 April–5 May 2002. The main activities of the Air Chemistry Group in the Hungarian Academy of Science, University Vezsprem are concentrated on the measurement and analyzes of the chemical composition of aerosol. The mass concentrations of inorganic ions, water-soluble organic carbon, water-insoluble organic carbon and black carbon in atmospheric aerosol collected at three European background sites: (i) the Jungfraujoch, Switzerland (high-alpine, PM2.5 aerosol); (ii) K-puszta, Hungary (rural, PM1.0 aerosol); (iii) Mace Head, Ireland (marine, total particulate matter) are determined in collaboration with the Paul Scherer Institute, Villigen, Switzerland. A box model is applied to estimate the direct climate forcing of aerosol particles for rural air in Central Europe during summertime. Data from satellite observations and other surface measurements are used as input parameters for the model. Rural fine aerosol collected in Hungary is analyzed with different techniques to estimate the average molecular weight (AMW) of humiclike substances (HULIS). Air Pollution Modelling and Other Topics The capital of Hungary, Budapest, is the largest city in the 10 countries. The occurrence of winter and summer high air pollution episodes is typical for the city. Therefore, the main effort in Hungary is oriented towards air pollution problems in Budapest. The founders of air chemistry research and air pollution modelling in Hungary Prof. Meszaros and Dr. Szepesi are well known specialists in the international air pollution community. National standard models (Gaussian types) were developed in the seventies. They are routinely used for regulatory purposes in Hungarian towns. Several foreign models were implemented in Hungary in the last decade (US EPA ISC3, Norwegian model KILDER, etc). Hungary has participated in several international modelling projects, e.g. EUROTRAC, COST CITAIR, COST 710 and COST 715. The Hungarian Meteorological Service is a consortium partner of PHARE Topic Link on Air Quality. The main objectives of the current air pollution research in Budapest are: to investigate the urban photochemical processes inside the city and in the urban plume, the origin of toxic elements in atmospheric aerosols and to improve ozone prediction for Budapest. Combination of a statistical model and a causal model (US OZIPR) is being tested. The Aerosol Association of Hungary is a part of the European Aerosol Federation. Latvia Ambient Air Quality Monitoring Network The Ministry of Environmental Protection and Regional Development is responsible for the co-ordination of environmental monitoring in Latvia and the Latvian Hydrometeorological Agency (LHMA) for implementation of the governmental policy in environment quality (water, air, and precipitation). The Latvian Environmental Consulting and Monitoring Centre (LECMC) as reference centre for monitoring activities, and the Latvian State Statistics Bureau (LSSB) as reference centre for environmental statistics are involved in

Survey on Atmospheric Chemistry Research in Some New EU Member States

313

air quality monitoring (Latvia, 1999). The first two institutions are subordinated to the Ministry of Environmental Protection and Regional Development. There is one National Air Quality Network in Latvia. During 1997 air monitoring systems running semi-manual air sampling (at 8 posts, 6 cities) were replaced by automatic air quality OPSIS DOAS stations. Since 1998 the network has consisted of six stationary city stations in three towns and two villages. Two stationary city stations (Riga) are included in the EUROAIRNET. The background air quality monitoring includes two stations. Station Rucava was established in 1985, station Zoseni in 1994. SO2, NO2, major ions, heavy metals and other components are measured at both stations; O3 is measured at one station. Both stations are involved in international activities (regional GAW/EMEP/ICP-IM) and are included also in the EUROAIRNET. On-site data are collected on paper charts. Data are available in the form of paper sheets and reports from 1986 (Background Network) and since 1969 (National Air Quality Observation Network - NAQON). From 1991 the data are available in electronic form. Since 1998 raw data from NAQON (excluding background stations) are received via modem line at LHMA. For data analysis the reporter of the OPSIS "EnviMan" management system is used. There is no uniform central database for all the measured data in Latvia (Latvia, 1999). For EUROAIRNET, Latvia has selected two stations in urban areas (in the capital of Latvia – Riga) and two stations have been selected in rural areas. The information is submitted through the Data Exchange Module. The site classification of selected stations and the Quality Assurance and Quality Control (QA/QC) issues have been elucidated. The chemical analyses of manual samples are performed in LHMA's laboratory, which passed accreditation and received certification in January 1998. The calibration of DOAS OPSIS monitors is performed on a regular basis. Regarding the establishment of an ambient air monitoring system in Latvia (Latvia, 2000) compliant with EC directives a number of EU regulations were adopted, namely, the Council Directive 96/62/EC of 26 September 1996 on ambient air quality assessment and management, the Council Directive 92/72/EEC on air pollution by ozone (Information and data exchange Formats (April 96)), the Council Directive 99/30/EC of 22 April 1999 relating to limit values for sulphur dioxide, nitrogen dioxide and oxides of nitrogen, particulate matter and lead in ambient air, and the Council Decision 97/101/EC of 27 January 1997 establishing a reciprocal exchange of information and data from networks and individual stations measuring ambient air pollution within the Member States. Atmospheric Chemistry Research Latvian tradition in chemistry research goes back in the 19th and 20th century with Wilhelm Ostwald born in Riga in 1853, who was awarded the Nobel Prize for Chemistry in 1909. Ostwald and Arrhenius (who worked in Riga in 1886) became founders of classical physical chemistry.

314

E. Batchvarova et al.

The Institute of Atomic Physics and Spectroscopy (IAPS), University of Latvia has planned a contribution to ACCENT-TROPOSAT-2, Task Group 3 for the period 2004-2008. Within the framework of the TROPOSAT programme the Institute will continue atmosphere chemistry research through laboratory and field experiments (Starnet, 2005). A MAXDOAS Ground validation station for the GOME (Global Ozone Monitoring Experiment) will be built, including a SCIAMACHY (Scanning Imaging Absorption spectrometer for Atmospheric CHartographY). Intense cooperation with colleagues from the University of Bremen is expected. The participation of the institute in laboratory experiments of troposphere and stratosphere chemistry is also a part of ACCENT-TROPOSAT-2 activities. Funding agencies might be the Latvian Science Council, the Ministry of Education and Science, University of Latvia, EU Structural funds, Other EU projects. Within the EUROTRAC-2 Subproject TRAP45, a team from IAPS has analyzed little used data sources on the air pollution history of Lativia. The influence of the transition from a centralized and unified administration under the Soviet Union to a national administration on air quality is elucidated in Übelis and Bergmanis (2000). Studies on size distribution and composition of airborne particles are performed at the University of Latvia, Department of Analytical Chemistry. Air Pollution Modelling and Other Topics The former USSR standard OND-86 (description of the model was included in ETC AQ Model Documentation System) is still in use in Latvia for regulatory purposes. The Swedish model system AIRVIRO was implemented in Riga in 1996 and also in Ventspils. No national air pollution models are being developed in Latvia at the present time. The advanced air pollution monitoring system OPSIS (14 monitoring stations, including 4 stations in Riga) has recently commenced operations. All the monitoring stations are connected to a reciprocal information exchange network and are being gradually connected to EUROAIRNET. Lithuania Ambient Air Quality Monitoring Network The air quality monitoring in Lithuania is managed by the Ministry of Environment (MoE) and is a sub-program of the National Monitoring Programme (NMP) (Lithuania 2001). For most tasks connected with NMP's coordination, the Joint Research Centre (JRC) performs management, data processing and reporting. JRC together with 8 Regional Departments (RD) of MoE deals with monitoring air quality, air emission, and surface water and ground water quality. Ambient air quality monitoring is carried out in 5 RDs, where stationary AQ control stations are located. Regional Departments are responsible for sampling and analyses in their local areas. JRC is responsible for data gathering, data processing and reporting. One part of the air quality monitoring, the background stations, are under the control of the Institute of Physics, which acts in this field as a subcontractor of JRC. Based on the type of impact, regional and local air pollution levels are discernible in Lithuania. Regional air monitoring is carried out at sites remote from industrial enterprises

Survey on Atmospheric Chemistry Research in Some New EU Member States

315

(Aukstaitija, Zemaitija, Dzukija, Preila) and is considered as a part of integrated monitoring. Aukstaitija and Dzukija National Parks stations started operation in 1993 (however, in the middle of 1999 the Dzukija station was closed), Zemaitija National Park in 1995. The monitoring station in Preila was opened in 1980 and is operating under international EMEP and EUROTARC programmes. All these stations allow the control of long-range trans-boundary air pollution brought to Lithuania from neighbouring and distant countries. Sampling in these stations is conducted daily (duration 24 hours) or weekly employing the method of air filtration using fibre filters. The measurement programme covers: NO2, SO2, SO4, NO3+HNO3, NH4+NH3, O3. Local monitoring is conducted in zones where point pollution sources or their complexes (cities and industrial centres) are significant, i.e. where the anthropogenic impact both on nature and human health is the most prominent. Air monitoring activity was initiated in Lithuania in 1967 and is basically restricted to a local level. Up to 2000 it consisted of 26 stationary air quality control stations (3 of them automatic) located in the biggest cities and industrial centres. Sampling of the main pollutants is carried out for 30 minutes three times per day (at 7 a.m., 1 p.m. and 7 p.m.). Main measured components are: TSP, SO2, NO2, CO, SO4, NO, H2S, NH3, HF, CH2O (formaldehyde) and phenol. Ozone and PM10 are established in automatic stations, located in Vilnius, as well as SO2, NO2, NO, CO. According to the Air Monitoring Strategy adopted by the MoE authorities and requirements of EU Directives, a new Lithuanian local air-monitoring network is being developed. After the implementation issues listed in the strategy and installation of equipment (supported by PHARE project), there are 5 stations in Vilnius, 1 in Kaunas, 2 in Klaipeda and 1 in Siauliai, Panevezys, Mazeikiai, Kedainiai, Jonava, and N.Akmene. The requirements of EU Council Directive 1999/30/EC for limit values of SO2, NOx, PM10 and lead have been adopted in the statement law of Lithuania as well as the requirements of the Framework Directive. The sampling procedures and the analytical methods are unified through the whole monitoring network. All chemical laboratories perform internal and external inter-calibrations. Quality Assurance procedures apply in these laboratories according to the national standards and norms. Regional Department laboratories carry out internal control based on a programme performed by JRC. JRC organises external control for all RD laboratories a minimum 2 times per year. All controls include the use of calibration standards, calibration curves, check of precision, frequent processing of blank samples, replicate analyses. Accreditation has not yet been implemented. All raw data are collected in the laboratories on paper and in electronic format. The Regional Departments and the Institute of Physics (which is responsible for background station data) send the data in digital form every month to JRC, where the data are processed and stored. Quarterly reports "State of Environment in Lithuania" and "Air Quality Annual Report" are issued. Information from specialised structures of the EoM (including Regional Departments, JRC, and Divisions of EoM) is supplied for different organisations and institutions. Atmospheric Chemistry Research The Institute of Physics is a state scientific research institute of Lithuania. It merges scientists and laboratories for basic and applied research in chemical physics and biophysics,

316

E. Batchvarova et al.

modern laser spectroscopy and nonlinear optics, nuclear physics, physics of atmosphere and some other related areas of physics. In addition to the scientific research, the Institute is engaged in educational activities via stringent connection links with several universities of Lithuania. The researchers from the Institute are expected to give professional advice on issues connected with environmental pollution, nuclear energy, etc. In some of the mentioned activities the Institute of Physics is the main or even the only research institution in Lithuania. The Institute of Physics formerly belonged to the Lithuanian Academy of Sciences. Due to reforms in the scientific research organisations carried out in 1991, when the independence of Lithuania was restored, the status of the Institute was changed. The transformations of pollutants in the atmosphere are investigated at the Institute of Physics, Environmental Physics and Chemistry Laboratory (EPCL, IP). Such studies are needed, as the development of cost-effective strategies for the control of pollutant emissions requires accurate assessment of deposition fluxes. It is important to understand the mechanisms of predominant chemical reactions in the atmosphere, to evaluate the rates of gas-to-particle conversion, and to include the information obtained in parameterised form in deposition models. Information about mechanisms of chemical reactions predominant in gas-to-particle conversion processes is obtained by comparing the experimentally determined data with the theoretically predicted particle diameter growth. It has been shown that during summer-months condensation of low-pressure vapour formed by gas phase homogeneous chemical reactions is a predominant mechanism of the continental aerosol particle growth. In winter, the growth of a sub micrometer continental aerosol particle is frequently governed by two competing processes: condensation of low-pressure vapour and heterogeneous oxidation of aerosol precursor gases inside the liquid droplet. Ozone is found to be an active oxidant in the heterogeneous aqueous phase chemical reactions. A number of atmospheric chemistry studies (Long-range transport of gases and particles; cycles of sulphuric and nitrogen compounds in the atmosphere; dynamics of organic compounds in the environment atmosphere, hydrosphere, and soil; variations and changes of total ozone; mechanisms of gas-to-particle conversion in the atmosphere; hygroscopic properties of atmospheric aerosols; evaluation of dry and wet fluxes of atmospheric pollutants to various ecosystems; exchange of chemical and biological constituents at the air/sea interface; seasonal variations of composition and concentrations of airborne fungi in Lithuania; and modelling of the exceedences of critical loads of sulphur and nitrogen for various ecosystems) are performed at EPCL, IP. Extensive international cooperation takes place at the institute with the following major activities: EC project BASYS (Baltic Sea System Study); NATO project on Relationship between Air and Waterborne Micro-organisms in Baltic Waters with the University of Cincinnati, Cincinnati, USA; EUROTRAC-2, subprojects BIATEX-2 (Atmosphere/Biosphere Exchange of Pollutants), AEROSOL (Aerosol Balance in Europe), SATURN (Studying Atmospheric Pollution in Urban Areas), TOR-2 (Photochemical Oxidants over Europe), and MEPOP (Interdisciplinary Research on Mercury and POPs); WMO (World Meteorological Organization) GOOS (Global Ozone Observing System) programme; ICP/IM (International Co-operative Programme on Integrated Monitoring of Air Pollution Effects on Ecosystems) programme; EMEP (European Monitoring and Evaluation) programme; HELCOM (Baltic Marine Environment Protection,

Survey on Atmospheric Chemistry Research in Some New EU Member States

317

Helsinki Commission) programme; COST (European Cooperation on Scientific and Technical Research) programme (COST-319- Transport and Air Pollution). Air Pollution Modelling and Other Topics Detailed urban air pollution studies are continuously carried out for the two biggest Lithuanian cities, Vilnius and Kaunas. Both monitoring and modelling tools are applied. Until the year 1998 the only approved air pollution regulatory model in Lithuania was OND-86. The advanced Swedish modelling system AIRVIRO was installed in Vilnius in 1996, and in Kaunas in 2000. For routine regulatory purposes the US EPA models ISC3 and SCREEN3 were recently implemented. The main limitations of urban modelling in Lithuania are: unsatisfactory traffic input data, absence of upper air meteorological measurements and limited information about international air quality modelling projects. Advanced techniques linking numerical models and observation data are under development at The Institute of Physics, Atmospheric Pollution Research Laboratory (APRL, IP). These are used to determine the complex interactions between the sources, the concentrations and deposition fluxes of trace substances having impact on the environment. Recent work has concentrated on the study of the dynamics of gaseous (NH3, HNO3, SO2, NOX, O3, CO2, H2S, etc.) as well as particulate (SO42-, NO3-, Cl-, NH4+, Na+, Ca+, K+, B(a)P, etc) pollutants in the atmosphere and on the evaluation of biosphere/atmosphere exchange fluxes of photo-oxidants, acidifying substances, heavy metals and persistent organic pollutants. Furthermore, the migration of heavy metals and persistent organic pollutants in the terrestrial and aquatic ecosystems has been studied. The serious pollution episodes in the urban environment are not generally caused by the sudden increase in the emission of pollutants, but are rather a result of short-term unfavourable meteorological conditions. A cost-effective way to prevent the occurrence of high air pollution episodes might be the temporal reduction in emissions that is based on the meteorological and air pollution forecast. The main fields of research at the institute include: transformation, migration and bioaccumulation of trace metals in the environment; mercury cycling in the environment, identifying mercury sources and sinks; influence of meteorological parameters on the urban air pollution and its forecast; manufacture of portable Mercury Vapour Analyzers Model GARDIS-5; development of new sampling and analytical methods, developing automated Air Pollution and Meteorological Monitoring Systems. International collaboration is achieved through projects like BASYS (Baltic Sea System Study); EUROTRAC (European Experiment on the Transport and Transformation of Environmentally Relevant Trace Constituents in the Troposphere over Europe); EMEP (Cooperative Programme for the Monitoring and Evaluation of long range Air Pollutants in Europe); HELCOM (Baltic Marine Environment Protection Commission); ECOSLIT (Ecological Sustainability of Lithuania); NORD (Atmospheric Heavy Metal Deposition); collaboration with Lund University, Sweden; co-operation project with Stockholm University; research collaboration with Gothenburg University in Mercury compounds determination; collaboration with EPA in Chicago, USA; collaboration with Tronheim University, Norway.

318

E. Batchvarova et al.

Poland Ambient Air Quality Monitoring Network Air quality monitoring in Poland is performed by The State Inspectorate for Environmental Protection SIEP (within the structure of the Ministry of Environmental Protection, Natural Resources and Forestry) with its provincial (voivodship) branches VIEPs; The State Sanitary Inspectorate SSI (subordinated to the Ministry of Health and Social Welfare) with its provincial branches VSIs; industrial plants or companies and research institutes, foundations, etc. In accordance with the Statute of the State Inspectorate for Environmental Protection (SIEP), the Chief Inspector of SIEP plays the co-coordinative role within the State Environmental Monitoring Programme. The State Sanitary Inspectorate (SSI) is responsible for air quality assessment from the viewpoint of health and hygiene requirements. The authority responsible for official data on national emissions into air is the Department of Environmental Protection in the Ministry of Environmental Protection, Natural Resources and Forestry (MoE). Air quality monitoring data are used at national, voivodship (provincial) and local levels. At the national level, the General Monitoring Network of air quality (GMN), established in 1985, is the largest one. It is subordinated to the SSI and operated by its provincial (or local) branches. Its main purpose is to supply information on population exposure to air pollutants for the needs of the National Health Programme. Data are also used for environmental policy purposes. Stations belonging to the network are also a part of lower level monitoring systems - the collected data are used both at the local and voivodship level. In 1997 the network consisted of 506 stations, located in cities above 20 thousand inhabitants and in health resorts. The basic monitored pollutants are SO2, NO2, black smoke and TSP. In total, 37 pollutants (from 1 to 19 compounds at each station) were monitored in 1997. Almost exclusively manual 24-hr measurements are carried out at the stations. The network is supervised by the Nofer Institute of Occupational Medicine (since 1992 till 1999), which is responsible for QA/QC, data collection, assessment and reporting. National Network of Basic Stations (NNBS), established in 1991, is subordinated to the SIEP. The network consists of selected monitoring stations belonging to Voivodship Inspectorates for Environmental Protection (VIEPs), to SSI (i.e. selected stations of GMN), research institutes and industrial plants. Stations incorporated into the network also work (or supply data) for other monitoring systems (local, provincial, regional and international programmes). The NNBS is a regulatory and general assessment network for air pollution and its trends in Poland. In 1998 there were 95 monitoring stations in NNBS, located both in urban and rural areas. The measurement programme comprises 24-hr concentrations of SO2, NO2, suspended particles (PM10/TSP/BS) measured at almost all of the stations, and 30-min concentrations of ozone and traffic-related pollutants at some of the stations. In 2001 the NNBS consisted of 106 air quality monitoring stations, 95 urban and 11 rural, including 4 EMEP stations. Concentrations of SO2 and NO2 were measured at about 100 sites, O3 at 23 sites. The network is operated under the supervision of the Institute of Environmental Protection (IEP), which is responsible for data collection, assessment and reporting and, to some extent, for QA/QC. At the voivodship level the most comprehensive and broad set of monitoring data is used. The monitoring data (operated by different institutions) are

Survey on Atmospheric Chemistry Research in Some New EU Member States

319

collected in voivodship databases and are used by provincial authorities and Voivodship Inspectorates of Environmental Protection VIEPs. The Air Quality Monitoring (AQM) data are used mainly to support environmental policy and strategy planning at a voivodship level (which seems the crucial level from the point of view of environmental planning in Poland), and to assess its effectiveness. Regional AQM networks provide information on air quality in a given region to support environmental and economic policy making. The extent, density and measurement programmes of the networks depend on the air pollution problems in a given region. The most advanced, fully automated, regional networks exist in the Black Triangle region and the Silesia agglomeration. Some regional networks have a warning function, providing up-to date information on existing and/or predicted levels of pollutants dangerous to health (e.g. the Automated Network in the Silesia agglomeration). Data from the regional networks are used at the voivodship and local level. Some of the stations are included in the national networks. Local AQM networks cover the area of a city or the area potentially influenced by a specific industrial plant. The networks are established by or with participation of local governments. In the case of the networks owned by industrial plants, the networks are established and operated by the plants/companies on the basis of administrative decisions. Their task is to provide information on the local air quality, local compliance with pollution standards and to evaluate pollution abatement programmes prepared by the industrial plants. In many cases the local networks are planned to serve emergency/warning purposes. Data from selected stations belonging to the local networks are used for air pollution assessment at a higher level. Again, the extent, density and measurement programmes of the local networks depend on the air pollution problems in a given city or the air pollution caused by the industrial plants. In 1998 there were about a dozen automated local networks, 8 of them operated by industrial plants/companies. Poland participates in a number of international programmes like GAW/WMO, HELCOM, EMEP and GEMS/AIR. The monitoring stations (MSs) included in the international programmes are located over the whole area of Poland and are elements of the national monitoring networks. The EMEP, GAW/WMO network includes four background stations. One of these MSs works additionally also for HELCOM. All of the stations belong to NNBS. Six urban stations (3 in Warsaw, 3 in Wroclaw) operated by the State Sanitary Inspection (SSI) and belonging to GMN contribute to the GEMS/AIR international Programme. NNBS prepares monthly and yearly reports for the SIEP. GMN prepares yearly reports. At the Voivodship level daily reports for mass media, yearly (in some voivodships daily and weekly and monthly) reports are released. The local networks data are used by industrial managers and made available to the local authorities and Voivodship Inspectorates for Environmental Protection (VIEP), and in some cases presented to the public. At the EMEP, GAW/WMO stations the standards and recommendations concerning data reporting are kept. Data are sent to the co-ordinating centre at NILU in Norway. At the GEMS/AIR stations the data reports are sent to the GEMS/AIR Centre in Geneva once a year. At all stations and laboratories the QA/QC procedures are systematically implemented, such as EU standards, inter-calibration, external audits, validation and verification. At EMEP,

320

E. Batchvarova et al.

GAW/WMO and GEMS/AIR stations the standards and recommendations of the programmes are kept. Basic measured parameters are SO2, NO2, NOx, and PM10/TSP/BS. Additionally, O3, CO, VOC, THC, F2, C12, HCl, NH3, phenol, HCHO, H2S, PAH, PAN, meteo data, turbidity, Cd and Pb in dust fall (monthly samples), and other heavy metals in 24-hour samples are measured at some stations. Atmospheric Chemistry Research The Central Chemical Laboratory (CLCh) is the primary analytical facility of the Polish Geological Institute (PIG) and carries out practically all the chemical analysis at PIG. CLCh took part in the INCO-COPERNICUS Project: “Development of Analytical Procedures to Guarantee Quality Assurance in International Environmental Monitoring”. CLCh participated in numerous international inter-laboratory comparisons as well as organized various inter-laboratory comparatives and proficiency testing studies for laboratories of Provincial Environment Protection Inspectorates. In March 2000, CLCh was granted the Certificate of Testing Laboratory Accreditation, with respect to physical and chemical analyses of water. The Institute for Meteorology and Water Management (IMGW) was founded in 1919. In 1972 IMGW was re-established by merging the Polish Institute of Hydrology and Meteorology (PIHM) with Institute for Water Management (IGW). IMGW is the responsibility of the Ministry of Environment. It has several Regional Branches in Bialystok, Kraków, Gdynia, Katowice, Poznan, Slupsk, Warsaw, and Wroclaw. The Institute has managed scientific and research studies in many fields related to the meteorological, hydrological and environmental problems, including atmospheric chemistry. Two laboratories of IMGW perform analytical measurements for natural water and air. The Laboratory of Atmospheric Chemistry performs analytical measurements for rainwater and precipitations. It participates in inter-comparison of analytical methods within EMEP and other programmes. The Institute for Environmental Protection (IOS) was founded in 1973. In April 1986, it was reactivated as an independent State Unit for scientific research operating under its own statute and subordinated to the Minister of Environmental Protection. Apart from the main centre in Warsaw, the Institute of Environmental Protection has two regional branches in Gdansk and Wroclaw. The Institute has a special entitlement from the Ministry of Environmental Protection to provide expert opinions. The Institute has chemical and biological laboratories, which provide the chemical and biological data. At the Institute of Environmental Engineering (IEE), Zabrze, Polish Academy of Sciences (IEE PAN) research is performed on: mechanics and physico-chemistry of pollutant generation and identification of its origin; transport and transformation of pollutants in the environment; methods of particulate matter and gaseous air pollutants' emission control; studies of environmental contamination and many other topics. The scientific equipment at IEE is of a high standard and modern. The Department of Air Protection at IEE deals with: theory and technology fundamentals of the removal of SO2 and NOx from flue gases; kinetics of homogenous and heterogeneous reactions; study and modelling of the emission from low and unorganized industrial and transport

Survey on Atmospheric Chemistry Research in Some New EU Member States

321

sources with methods of FTIR – spectroscopy; application of spectroscopic methods for the study of the air condition in the surroundings of industrial and transport sources; theoretical fundaments and management instruments of the air quality in urbanized areas etc. The Laboratory of Applied Analytical Chemistry at the Warsaw University, Chemistry Department works on environmental analytical chemistry, development of analytical methods for determination of trace metals in environmental samples and speciation analysis as a modern tool for environmental risk assessment. User groups of the EUROCHAMP infrastructure from Poland are the Technical University Wroclaw with the investigation of atmospheric chemical processes, automatic manual and passive monitoring, temporal and spatial modelling of air pollution with advanced statistical tools and the Institute of Meteorology & Water Management, Wroclaw with ambient air monitoring of air pollutants, aerosols, ozone, radioactive compounds as well as background air pollution monitoring (according to EMEP standards), measurements of pollutant loads incoming with precipitation. The Institute for Ecology of Industrial Areas (IETU) has two research (monitoring) stations. The Katowice Research Station operates within the National Atmospheric Pollution Monitoring Network which is part of the National Environmental Monitoring System. Since 1989 the station conducts regular monitoring of air pollution and precipitation quality. In 1999 it was incorporated into the European Monitoring Network of Air Quality EUROAIRNET of the European Environmental Agency. The Brenna Station has been established as a unit of the integrated environmental monitoring, performed within the framework of the Geneva Convention on Long-Range Trans-boundary Air Pollution. At present, the station is operated as an experimental facility to conduct studies on the impact of elevated ozone and UV-B concentrations on the genetic structure of plant populations. IETU’s research and development activities are aimed at formulating the scientific basis for policies and strategies addressing the protection of the natural environment of industrialised and urbanised areas. The main projects are related to characterisation of sites contaminated with organic and inorganic pollutants, impact of pollutant re-suspension on the air quality, secondary air pollution as a source of human exposure, monitoring of air pollution, etc. Among the research areas of the Institute of Physical Chemistry (ICHF), Polish Academy of Sciences (PAS) (ICHF, 2005) are photophysics and photochemistry of organic molecules, involving spectrally and time resolved laser techniques and kinetics and mechanism of chain reactions in homogeneous/heterogeneous systems, as related to sulphur dioxide transformations in the environment: physico-chemical fundamentals of desulphurization processes; atmosphere chemistry including impacts of organic and inorganic scavengers of sulphoxy radicals; interaction between these radicals and bioorganic compounds. As a contribution to a subproject of CMD of EUROTRAC-2, scientists from ICHF studied experimentally the effect of D-pinene and cis-verbenol on the rate of S(IV) oxidation catalyzed by Fe, as well as the ozone-affected autoxidation of aqueous SO2 in the presence of calcium. The investigations were aimed at clarifying the mechanism of the aqueous phase oxidation of SO(IV) simultaneously by O3 and O2, calcium acting as condensation nuclei. Inhibition of the S(IV) autoxidation in the atmosphere by secondary terpenic compounds was also studied

322

E. Batchvarova et al.

experimentally. A method was proposed for quantification of the inhibiting effect. All laboratory experiments at ICHF were carried out in a batch, perfectly mixed reactor, which had no gas space, in order to investigate the degradation of isoprene in the presence of sulphoxy radical anions. The aqueous-phase oxidation of isoprene can produce secondary pollutants, and influence transformation and the long-range transport of SO2 in the atmosphere. A team of scientists from the Chemical Faculty, Technical University of Gdansk, Poland and the Department of Chemistry, University of Waterloo, Canada presented monitoring results and an environmental pollution assessment for the Gdask-Sopot-Gdynia Tricity (Poland), based on an analysis of precipitation. Precipitation samples were collected over a period of 12 months (January–December 1998) at ten locations in the Tricity. An attempt was made to explain the co-occurrences of certain ions and the significance of their mutual effects. The use of magnetometry as a fast and cost-effective proxy method for screening and monitoring of anthropogenic pollution over central Europe is proposed within the FP5 MAGPROX (MAGnetic PROXies) project (2000-2003). The project is focused on a large area affected by industrial activities, spanning Germany, Austria, Czech Republic and Poland. The method can be also used for high resolution mapping of pre-selected sites showing anomalous geochemical patterns in order to allow a proper sampling for obtaining additional information. The Air-Sea Interaction Laboratory, Institute of Oceanology, PAS, Sopot conducts research on: mass, energy, momentum and radiation fluxes across the sea surface; the concentration and size distribution of marine aerosol in the boundary layer over the sea surface and in coastal zones; air-sea interaction phenomena, including sea-atmosphere chemistry. Research into aerosol production and its properties in the marine boundary layer was made during the AREX campaigns of 2000-2003 in the open Baltic Sea. Results of cloud water and air chemistry field measurements at Mt. Szrenica in Poland have been reported by scientists from Brandenburg Technical University Cottbus and the Technical University Wroclaw, Institute of Environmental Protection Engineering. Aircraft, large and small balloons, ground-based instruments and satellites are being used to measure ozone and other atmospheric gases and particles. The combined activities aim to improve the understanding of Arctic ozone depletion and upgrade satellite observations of the ozone layer. The joint initiative involves over 400 scientists from the European Union, Canada, Iceland, Japan, Norway, Poland, Russia, Switzerland and the United States. Several institutions in Poland participated in human health-oriented international projects related to atmospheric chemistry. The Institute of Occupational Medicine and Environmental Health (IOMEH), Sosnowiec analyzed measurements of PM10 and PM2.5 at Warsaw and their impact on human health within the frame of the CESAR study (1995-1996). The (National Institute of Hygiene (NIH), Warsaw (as a contributing member of the APHEIS group) made an health impact assessment of air pollution in Krakow (2001). The Institute for Ecology of Industrial Areas (IETU), Katowice and IOMEH were involved in the EMECAP project, with main objective being the improvement of the tools for Regulatory Bodies to plan actions to safeguard citizen’s health from mercury pollution.

Survey on Atmospheric Chemistry Research in Some New EU Member States

323

Air Pollution Modelling and Other Topics Poland is the largest country among the 10. The origins of air pollution modelling may be traced to the seventies, when the first simple models, both statistical and Gaussian, were introduced. The Gaussian model is chosen by the Ministry of Environment as a guideline for dispersion calculations from point sources. Several improved versions of the Gaussian model are still widely used in Poland (e.g. EK100W from ATMOTERM Ltd. Opole, which is included in ETC/AQ MDS). From the beginning of the eighties the main effort was put into adaptation and development of numerical models. Currently, there are numerous institutions engaged in this research: universities, state research institutes, local government institutions and private companies. The basic limitations of such activities at present are: shortage of funds available for research and lack of co-ordination on a country scale. Some advanced foreign urban models were implemented in Poland. Several top Polish specialists are working in EMEP MSC-W in Norway (Bartnicki, Olendrzynski, Pacyna). Only two years ago the first mesoscale meteorological prognostic model (Unified Model from the UK Meteorology Office) was put into operation at the Interdisciplinary Centre for Computer and Mathematical Modelling of Warsaw University. Also, at the Institute of Meteorology and Water Management the newest non-hydrostatic mesoscale model from the German Weather Service is being implemented. Poland´s capital city, Warsaw, is situated in flat terrain and this, combined with a high average wind speed leads to relatively good dispersion conditions. The air quality data from Warsaw suggest that maximum sulphur dioxide city background levels do not exceed WHO-AQG. The meteorological summer smog potential is also not high. The improvements of urban air quality in Poland can be illustrated by the city Krakow, where a three-fold decrease in sulphur dioxide and suspended particulate concentrations has been observed over the last decade. The new automatic monitoring devices were installed in 1991. The collection and analyses of data in a central database facilitates forecasting of pollution episodes based on the USA EPA modelling system. Romania Ambient Air Quality Monitoring Network In Romania, the monitoring of air quality is organized under the authority of the Ministry of Waters, Forests and Environmental Protection (MWFEP) and the Ministry of Health. There are four mobile laboratories for air quality measurements, three of them at units subordinated to MWFEP (Bucharest Environmental Protection Agency, Bacau Environmental Protection Agency, Institute for Research and Environmental Engineering) and one at Bucharest Town Hall. The mobile laboratories are equipped with automatic analysers. The National Air Quality Network, subordinated to MWFEP, is one of the subsystems of the National Environmental Monitoring System. It is developed according to the Master Plan elaborated within the PHARE Programme. Air quality monitoring carried out at the international level is concentrated in background pollution stations as part of BAPMoN and EMEP networks. Also the local trans-boundary pollution along the Danube River is monitored in several stations. At the local level, common pollutants (SO2, NO2, NH3, suspended powders) are measured at all stations and specific

324

E. Batchvarova et al.

pollutants (HCl, phenols, aldehydes, Cl2, H2S, CS2, F, H2SO4, heavy metals: Pb, Cd) in industrial zones in relation with different activities. At most stations manual 24-hour sampling is applied. At a few sites, semi-automatic systems with 30 minutes sampling at 3-hour intervals are installed. The collection of sediment particles is performed via monthly sampling. The analysis of gases is based on chemical methods, spectrophotometry UV-VIS; of particles on gravimetry; of heavy metals on atomic absorption spectrophotometry. The National Network for Precipitation Quality Monitoring includes approximately 90 stations located in rural and urban zones. A brief description of this network includes 3 main items: sampling – weekly; analysis - pH, conductivity, acidity/alkalinity, concentrations of the main anions and cations (Na+, K+, Mg2+, SO42-, NO3-, Cl-) currently at the background stations and soon in the whole national network; and methods - spectrophotometry. The inter comparison and calibration procedures of Air Quality Monitoring in Romania are strictly followed. Atmospheric Chemistry Research Within the frame of the subproject GPP of EUROTRAC-2, scientists from the Faculty of Chemistry, Al. I. Cuza University Iasi (FC,UI) together with scientists from the Bergische University Wuppertal, Germany studied experimentally the OH -initiated oxidation of dimethyl sulphide (DMS) at sub-zero temperatures using a quartz glass reactor. Chemical transformations of DMS due to its reactions with OH radicals during the day and NO3 radicals during the night lead to the formation of sulphur containing species, both inorganic and organic in nature, which may significantly contribute to the acidity of the atmosphere. Under Romanian-German-Spanish collaboration, experimental investigations of the gasphase reactions of the NO3 radical with a series of benzenediol compounds were performed. The experiments were carried out in two chamber systems with in situ FT-IR (Fourier Transform – Infrared Spectroscopy) detection of reactants: a 1080 l quartz glass reactor at the Bergische University Wuppertal and in the EUPHORE outdoor smog chamber facility in Valencia/Spain. The kinetics of the reaction of NO3 radicals with three benzenediols using a relative kinetic technique have been investigated. As a contribution to the subproject CMD-GPP of EUROTRAC-2, experimental studies on the kinetics, products and aerosol formation from the reaction of O3 with 3 benzenediols were carried out in a smog chamber under controlled conditions. Within the frame of the EU EXACT project (Romania, Germany, France and Spain), studies of the mechanisms of Secondary Organic Aerosol (SOA) formation from the photooxidation of aromatic hydrocarbon systems were performed. Experiments were carried out in the EUPHORE chamber facility. Under collaboration between FC, UI and the Bergische University Wuppertal, experimental studies of the atmospheric chemistry of C3 to C5 alkyl iodides have been performed in the EUPHORE outdoor chamber in Valencia, Spain. The kinetics and mechanism of the reactions of O atoms with alkyl halides have also been investigated.

Survey on Atmospheric Chemistry Research in Some New EU Member States

325

A contribution to the subproject Chemical Mechanisms Development - Gas Phase Processes (CMD-GPP) of EUROTRAC-2 was also the study on “Development of Oxidation Mechanisms for Aromatic Hydrocarbons and their Unsaturated Difunctional Products”, made with German and Japanese (National Institute for Environmental Studies, Atmospheric Environment Division, Tsukuba) scientists. All the above mentioned activities are the result of fruitful collaboration between the following institutions: Bergische Universität Wuppertal, Germany; "Al.I.Cuza" University of Iasi, Department of Analytical Chemistry, Faculty of Chemistry, Romania; Ecole des Mines de Douai, Département Chimie et Environnement, France; Fundación Centro de Estudios Ambientales del Mediterráneo, Paterna, Valencia, Spain; National Institute for Environmental Studies, Atmospheric Environment Division, Tsukuba, Ibaraki, Japan. The University of Iasi participates in the user groups of the EUROCHAMP infrastructure with the following fields of excellence: Mechanistic and kinetic indoor and outdoor photo-reactor studies in the gas-phase; Studies on the aerosol formation under simulated atmospheric condition using several reactions chamber facilities; Monitoring of air pollutants, aerosols. Many investigations made by Romanian scientists are dedicated to the interrelations between chemical reactions in the atmosphere and specific meteorological conditions. The National Institute of Meteorology and Hydrology, Bucharest is closely involved in the study of the effects of increasing the CO2 concentration on the extreme temperatures and drought episodes in Europe, as well as ozone depletion related studies. The International Association of Meteorology and Atmospheric Sciences (IAMAS) Organization for Romania, as a Section of the Romanian National IUGG Committee was constituted at the National Institute of Meteorology and Hydrology (NIMH), Bucharest, in cooperation with the Faculty of Physics, Department of Atmosphere Physics of the University of Bucharest. Many research laboratories in the field of air and water pollution monitoring are presently at the Institute of Environment Research and Engineering (IERE). Studies related to upper air are being performed especially at the Astronomical Institute and Romanian Civil Authority for Aeronautics (ROMATSA). The main activities of IAMAS include monitoring of the ozone layer, data processing and instrument maintenance. Long term monitoring has been carried out in Bucharest since January 1980 using a Dobson spectrophotometer. The research activities are focused mainly on ozone climatology related with meteorological conditions on a local scale, total ozone with respect to changes in atmospheric circulation, evolution of the ozone layer on a regional scale. Some attempts of total ozone forecast are made based on improvement of the forecast model ALADIN. The Bucharest station (part of the Global Atmospheric Watch (GAW) Ozone Network as "associated station") fulfils the interaction with other programs and activities. Romania is a permanent participant in the WMO/GAW - GOOS and some European Union Programmes concerning the monitoring of the ozone layer. The research orientation of IAMAS is also related to a development (in the NIMH) of an operational Air Pollution Forecast System. Its goal is to produce 3-day air pollution forecasts of the most important air pollution species on different

326

E. Batchvarova et al.

scales. In NIMH, “A pilot system for Urban Environmental Impact Assessment in relation with Urban Planned Land-Use, using Open-GIS technology and pollution levels estimation procedures - ASSURE (ASsessment System for URban Environment )” has been developed. Between November 1995 and October 1996, particulate matter concentrations (PM10 and PM2.5) were measured in 25 study areas in six Central and Eastern European countries: Bulgaria, Czech Republic, Hungary, Poland, Romania and Slovak Republic (CESAR study under PHARE). As a contributing member of the APHEIS group, Romania (Institute of Public Health, Bucharest, Romania) participated in the study concerning the health impact assessment of air pollution in 26 cities (Bucharest) in 12 European countries during 2001. Six European countries are involved in the EMECAP project (2001-2004): Italy, Norway, Slovenia, Sweden, Poland and Romania (University of Bucharest – Bucharest). The objective of the EMECAP Project was to improve the tools at the disposal of Regulatory Bodies to plan actions to safeguard citizens’ health from mercury pollution. Other Romanian institutions partially related to the atmospheric chemistry research are the National Research and Development Institute for Environmental Protection, Bucuresti; Politehnica University of Timisoara; Faculty of Geology and Geophysics, Bucharest; Romanian Academy, State Institute of Physical Chemistry, Supramolecular Chemistry and Interphase Processes Laboratory, Bucharest; Institute of Atomic Physics, Romania; National Research Institute of Cryogenics and Isotopic Technologies; Institute of Environment Research and Engineering; National Institute of Research and Development for Nuclear Physics and Engineering "Horia Hulubei"; Institute of Agrometeorology and Environmental Analysis of Agriculture, FMA- Applied Met. Foundation; Polytechnical Institute in Bucharest. Air Pollution Modelling and Other Topics The leading institution in urban air pollution modelling in Romania is the National Institute of Meteorology and Hydrology and Water Management. Several urban dispersion experiments were completed during the eighties (Bucharest, Baia-Mare, Copsa-Mica, Hunedoara). The main goals of this effort were to test urban dispersion models, plume rise models and the methods for determination of dispersion categories. Currently a climatological dispersion model (CDM) and a Lagrangian puff model (INPUF-U) are used in urban modelling studies. Descriptions of both the models are already included in ETC/AQ MDS. A meteorological pre-processor of wind field and dispersion parameters (OML scheme) is available. The French limited area prognostic model ALADIN version SELAM putting in operation. The capital of Romania, Bucharest, is situated on a flat plain, which will favour dispersion, but this is countered by the generally low wind speed (average wind speed below 2 ms-1).

Survey on Atmospheric Chemistry Research in Some New EU Member States

327

Slovak Republic Ambient Air Quality Monitoring Network The Hydrometeorological Institute (SHMI) is the main organization having responsibility with respect to air pollution monitoring in Slovakia (SHMI, 2004). The air monitoring system consists of 35 automatic stations in the whole country with measurements of SO2, NOx, PM10, PM2,5, O3, CO, benzene, H2S. At 22 stations heavy metals Pb, Cd, Ni, and As are measured manually. Additionally, there are 10 monitoring stations of different industrial manufacturers and 18 ozone stations. Monitoring of the regional air and precipitation quality in the Slovak Republic has been performed at SHMI since 1977. The first station, Chopok, was established in 1977. This station was included from the beginning in two international programs, GAW WMO and EMEP UN ECE. Up to 1993 a network of seven regional stations was gradually made operational. At present, the regional network consists of five stations, all are EMEP stations: Chopok, Topo níky, Liesek, Stará Lesná and Starina. The automatic urban air quality monitoring network in the Slovak Republic was put into operation in 1992/1993. At the same time the Slovak regional stations (EMEP) started to be equipped with ozone analysers. Sampling of volatile organic compounds, VOCs C2-C6, or the socalled light hydrocarbons, was started in autumn 1994 at the Starina station. The VOC measurements in Slovakia were implemented under the assistance of the Norwegian Institute for Air Research (NILU). The Starina station is one of the small numbers of European stations, included into EMEP network with regular sampling of volatile organic compounds. They are measured and assessed according to the EMEP method elaborated by NILU. Measurements of identical samples carried out in the Slovak Hydrometeorological Institute and in NILU showed a high degree of agreement. The Slovak Hydrometeorological Institute participated in the AMOHA (Accurate Measurements of Hydrocarbons in the Atmosphere) project. This project was carried out under the leadership of the National Physical Laboratory in United Kingdom and the Fraunhofer Institute in Germany. The aim of this project was to elaborate the European Directive for optimum sampling, measurement and evaluation of hydrocarbons in ambient air. The annual amount of precipitation depends on location and altitude. Concentrations of all chemical substances dissolved in rainwater are spatially rather conservative. Chemical composition of precipitation over the industrial continents is controlled by regional in-cloud processes. Below-cloud scavenging of gases is an ineffective process in most cases. It means that concentrations should reflect the large-scale distribution and trends of emissions. Therefore, pronounced horizontal and vertical concentration gradients should not be expected. It can be noted that the gradual replacement of bulk with wet only samplers in Slovakia caused problems in the trend analysis. The only significant decrease of sulphate concentrations and corresponding increase of pH values were observed at all Slovak regional stations. A QA/QC programme has been partly implemented, but further improvement is needed. Accreditation of network and laboratories is ongoing. Development of standard operation procedures is in progress. Unfortunately present budgetary and capacity limitations do not allow

328

E. Batchvarova et al.

its full implementation. The quality of analysis in the laboratory is very good and tested annually via GAW and EMEP inter-comparison measurements. The EMEP model overestimates the measurements for the territory of Slovakia, possibly due to the very complex topography. The correlation of calculated and measured concentrations decreases with altitude. In 1998, there were 7 monitoring stations for regional air pollution monitoring, characterized as pollution of a rural type. The measurement programme includes SO2, NO2 and heavy metal concentrations in atmospheric aerosol PM, Pb, Mn, Cu, Cd, Zn, Ni, V, Cr. Within the framework of the project “Co-operation programme between Flanders and candidate member states in the central and eastern Europe” monitoring of air pollution and audit of Quality System was fulfilled (SHMI, 2004a). Atmospheric Chemistry Research The Department of Analytical Chemistry, Faculty of Natural Sciences, Comenius University performs a number of studies related to atmospheric chemistry (ANALYTIKA, 2004). Among the research projects on the topic funded by different grants are “Analysis and characterization of humic substances in air particulates sampled by filter media for air monitoring by both RP-HPLC and gel chromatography methods”, “New Methods of Air Sampling and Analysis of Chemical Compounds in Air” and “Trace and ultra trace analysis of pyrethoides in air samples by HPLC”. At the Departament of Environment Sciences, Faculty of Chemical Technology, Slovak Technical University, Laboratory of Air Protection Technology (CHTF, 2004), studies on air pollution transport and kinetics of coal combustion in the atmosphere of carbon dioxide are performed. Health effects of air pollution are studied at the Institute of Preventive and Clinical Medicine. Air Pollution Modelling and Other Topics The Czech Republic and the Slovak Republic have a common history of air pollution modelling. The Slovak Republic was established in January 1, 1993. The old guidelines used for regulatory purposes were replaced in 1994. The US EPA ISC model was adapted to Slovak conditions (model MODIM). Additionally, the Slovak model AUTOMOD is recommended by the Ministry of Environment as the standard model for traffic pollution studies. Both models were included in the ETC/AQ MDS. Extensive urban air pollution studies were conducted within the framework of the project PHARE EU/93/AIR22. The very complex terrain of Slovakia limits to a great extent the application of simple dispersion models. The other current air pollution problem in Slovakia is the occurrence of ozone episodes. Implementation of 3D photochemical transport and dispersion model is desirable. Recently, the French mesoscale prognostic model ALADINversion Slovakia was put into operation at the Slovak Hydrometeorological Institute in Bratislava. It increased the potential of air pollution modelling in the Slovak Republic. A

Survey on Atmospheric Chemistry Research in Some New EU Member States

329

comprehensive review on the status of the atmospheric chemistry modelling in the Czech Republic and Slovak Republic is given in Zavodski et al (2000). Slovenia Ambient Air Quality Monitoring Network At the Environment Agency of the Republic of Slovenia (ARSO), within the Water Quality Section of the Monitoring Office there is a Chemical Analysis Laboratory (CAL), which performs physical and chemical analyses on samples from the environment (water, precipitation and air) as part of the national and international monitoring of water and air quality. CAL performs analyses of precipitation and air samples as part of the EMEP international air quality monitoring and the GAW/WMO. The analyses involve daily samples of aerosols and gases, and sampling at the Iskrba EMEP station at Koþevska Reka. CAL performs physical and chemical analyses (pH, electrical conductivity, main anions and cations) of daily and weekly precipitation samples taken by the Agency at ten measuring points in Slovenia. In 2003 the CAL successfully completed the accreditation procedure for the area of chemical testing of samples from the environment (water and precipitation) in line with the SIST EN ISO/IEC 17025 standard. In 2001, a project, financed by EU PHARE program and competed by the Slovenian company AMES d.o.o., replaced the existing (since the 1980s) network with a new, up-to date air pollution monitoring system. The Environmental Agency of the Republic of Slovenia (in Slovenian - ARSO) is a body within the Ministry of the Environment, Spatial Planning and Energy (ARSO, 2005). Through its mission and vision, the Environment Agency of the Republic of Slovenia is in step with the European Environment Agency, which functions as a consortium of several countries on an expert level and covers five integral sectors: air and climate change, water, nature and biodiversity, the terrestrial environment, waste and the material cycle. Since environmental information is important for the more successful formulation of environmental policy and monitoring of its implementation, they are joining the European network for transferring environmental information and are setting up a complementary system on the national level, which will allow them to establish links within Slovenia and in international circles, and will ease the processing, analysis, synthesis and comparison of data. Apart from the stations in the network of the Environment Agency, there are other automatic air quality monitoring networks, owned by major Slovenian Power plants, factories and local communities. Most of them are built by AMES. Major networks of this type belong to the Thermal Power Plant Šoštanj (10 stations with meteorological and gases measurements, one station also gamma dose rate measurements, emission stations on three stacks, hydrological station on Paka river), the Thermal Power Plant Trbovlje (6 stations with meteorological and gases measurements, emission stations on the stack, hydrological station on Sava river), the Thermal Power Plant Brestanica (one station with meteorological and gases measurements, one meteorological station, emission stations on the stack, hydrological station), the City of Celje,

330

E. Batchvarova et al.

local community (one station with meteorological and gases measurements and two central units), the Cinkarna d.d. Celje (several small stations, SO2 and PM10 and meteorological measurements), and the Nuclear power plant Krško (4 meteorological stations, one has SO2 monitor added, all have gamma dose rate measurements, plus additional 9 stations only for gamma dose rate measurements). AMES d.o.o. is a private high technological and research company. Its fields of work are: meteorological sensors design, development and production; design and production of automatic measuring stations (ambient, emission, hydrological, meteorological, agrological, airport meteorology, radiological); design and production of automatic measuring systems for the environment; research in air pollution modelling (forecasting models based on neural networks, neural networks based solar radiation modelling, air pollution dispersion modelling). Atmospheric Chemistry Research Jozef Stefan Institute, Dept. of Environmental Sciences, Ljubljana, Slovenia (IJS, 2005). The Department of Environmental Sciences is an interdisciplinary group of researchers specializing in the fields of environmental analytical chemistry, radioecology, biogeochemical cycling and metabolism of trace elements in biological systems, bio monitoring, evaluation of nutritional values of food and feedstuffs, environmental modelling, and impact assessments. The Centre for Mass Spectrometry and management of an Ecological Laboratory with a mobile unit also form an important part of the work of the Department. At the division of Environmental Analytical Chemistry, some of the topics are: determination of persistent and/or toxic organic pollutants (aliphatic and aromatic hydrocarbons, chlorinated compounds, phenols, and pesticides) in different environmental matrixes (water samples, sediments, soil, biological materials); qualitative and quantitative analysis of gas samples by mass spectrometry, etc. The Environmental modelling and assessment group works on environmental modelling and assessment. Such modelling includes integrated mass transfer, migration and accumulation of pollutants in/through environmental media. Special emphasis is made on uncertainty and validity of the modelling. Other sides of the work concerns environmental impact assessment and development of expert systems and evaluation procedures for most complex environmental issues in the country, environmental policy development, and support of decision-making, as well as strategic environmental assessments. Among the user groups of the EUROCHAMP infrastructure is the National Institute of Chemistry, Ljubljana, Slovenia, with the main research field “Reactivity of aerosols under haze conditions and chemical characterisation of aerosol particles”. Participation of Slovenia in projects: x As a contributing member of the APHEIS group, Slovenia (Institute of Public Health, Ljubljana) participated in the study concerning the health impact assessment of air pollution in 26 cities (Ljubljana) in 12 European countries during 2001. x Six European countries are involved in the EMECAP project: Italy, Norway, Slovenia (Jozef Stefan Institute, Ljubljana), Sweden, Poland and Romania (April, 2001 - March, 2004). The objective of the EMECAP Project is to improve the tools of Regulatory Bodies to plan the actions to safeguard citizens’ health from mercury pollution.

Survey on Atmospheric Chemistry Research in Some New EU Member States

331

x As a contribution to the subproject TOR-2 of EUROTRAC-2, at the Faculty of Chemistry and Chemical Technology, University of Ljubljana, Hydrometeorological Institute, ECOSENSE, Ljubljana in cooperation with Fraunhofer Institute, Germany, an analysis of the ozone spring maximum at the Krvavec was performed. x The Project METROPOLIS covers 17 countries and brings together 38 participants from organisations and research institutes dealing with metrology and environmental monitoring in Europe. Leader from Slovenian part is the Jozef Stefan Institute. The METROPOLIS thematic network pursues three main objectives in environmental metrology: to improve the performance of environmental measurement systems and their harmonisation at EU level; to foster the dialogue between those who provide measurement methods and associated services, and the users of measurement results; to prepare the ground for further integration of research expertise and resources in environmental monitoring across Europe. x At the National Institute of Chemistry (NIC), in the frame of CMD subproject of EUROTRAC-2, experimental studies of the role of soluble constituents of atmospheric aerosols in the aqueous–phase autoxidation mechanisms of S(IV) was studied. The research focused on atmospheric water droplets (clouds, fog), where soluble constituents of atmospheric particles may be important in aqueous SO2 oxidation under non-photochemical conditions. In the frame of CMD project laboratory experiments in a semi-batch continuous stirred tank reactor under controlled conditions (T, air flow rate, stirring), were made in order to study the autoxidation of S(IV)-oxides catalyzed by transition metal ions (Fe(III), Fe(II), Co(II), Cu(II), Ni(II), Mn(II)). These studies were carried out at the National Institute of Chemistry. x At NIC, as a contribution to subproject AEROSOL, EUROTRAC-2, laboratory experiments with a continuous flow of some trace gases under controlled conditions, were performed in a reactor chamber. The preliminary results of SO2 oxidation in synthetic solutions containing various catalytically active metal ions were discussed. A laboratory model study of the role of aerosols composition in the process of formation of secondary acidic species from gaseous and particulate precursors was performed. x MERCYMS: An Integrated Approach to Assess the Mercury Cycling in the Mediterranean Basin x BIOMERCURY: Worldwide remediation of mercury hazards through biotechnology x Certification of trace elements: Several projects with JRC, IRMM- International Measurement Evaluation Programme; Geel, Belgium In the period 2004-2005 there are a number of bilateral projects between Slovenia and Austria: x Heterogeneous reactions of atmospheric aerosols under controlled experimental conditions typical for haze (Kemijski inštitut, Ljubljana); x Development of equations of state for humid gases based on speed-of-sound data (Univerza v Mariboru Fakulteta za strojništvo, Maribor); x Nitrate natural background assesment and pollution sources in groundwater (Univerza v Ljubljani Biotehniška fakulteta, Ljubljana); x Investigation of photochemical properties of molecules important for atmospheric chemistry, especially bromine and nitrogen containing compounds (Institut Jožef Stefan, Ljubljana).

332

E. Batchvarova et al. Air pollution modelling and other topics

Because of very complex terrain the application of simple dispersion models is very limited in Slovenia. Traffic pollution and the high level of surface ozone are the main current air pollution problems in the country. No official standard model for regulatory purposes has been accepted in Slovenia up to present. The US EPA model ISC3 is used for routine dispersion calculations from point sources. Some other imported models were tested in Slovenia but only on research basis. A neural network forecasting model was developed for the Sostanj thermal power plant. No urban air pollution studies are reported from Slovenia. Air pollution modelling is performed at the Jozef Stefan Institute, Dept. of Environmental Sciences, Ljubljana, Slovenia (IJS, 2005), AMES d.o.o. and the Hydrometeorological service. Comparative and General Comments, Identified Needs and Problems Air Pollution Monitoring All countries have adopted the EU environmental legislation. The manual stations are being rapidly changed to automatic firstly in Czech Republic, Poland and Hungary, then in Estonia, Latvia, Lithuania and Slovenia and lastly in the Slovak Republic, Romania and Bulgaria. In the different countries the inter-institutional communication problem exists to different extents after the economical changes and the enlargement of the EU. The Ministries of Health and Ministries of Environment keep separate networks, and participate in different international cooperation. In some countries (Czech and Slovak Republic) the Ministries of Environment have delegated the ambient air monitoring to the meteorological services, which has kept and further developed the expertise within the country. In other countries (Bulgaria, Slovenia) the Ministries of Environment enlarged their monitoring activities based on the PHARE programme and other European funding and reduced the monitoring programmes of the meteorological services. In this way, the groups with expertise and tradition suffered a lack of funding and the newly established units suffered a lack of expertise and tradition. There is no easy solution to such problems and they also exist in the Western European Countries as well. What the Ministries of Environment can do is to allocate greater funding for research based on monitoring data (as in Germany, for example). The environmental data have to be easier accessible for research, than they presently are (for example in Bulgaria). As governmental bodies, the Ministries of Environment are the competent authority for the majority of activities related to environmental issues such as trans-boundary air and water pollution, troposphere ozone and climate change, for instance. The main task of a ministry on these issues is therefore to elaborate and implement the environment policy. As it is nearly everywhere within the Member States of the European Union, the ministries carry out only a few studies and analysis by themselves, but a lot of routine work is allocated to different administrations or institutions. This approach was not followed in some countries of the 10 and the research and international collaboration on different environmental programmes suffered for a

Survey on Atmospheric Chemistry Research in Some New EU Member States

333

long time. Another problem with the ministries during the period of changes was hidden in the change of staff, which has lead to losses of expertise. Atmospheric Chemistry Research In all countries there are research institutes and departments of universities, where atmospheric chemistry research is performed. The groups from the Czech Republic, Poland, Hungary, Slovenia, Estonia and Lithuania are more active. The groups from Romania, Bulgaria and Latvia have worked mainly within international projects, using facilities in different European countries. In the countries where the scientific structure was not changed with the start of economic changes in 1990s, namely where the academies of sciences and their institutes on physics and chemistry were kept, one can see the most successful groups (Aerosol Laboratory, Institute of Chemical Process Fundamentals, Prague, Academy of Sciences of the Czech Republic for example). These groups kept the expertise and were able to join and substantially contribute to European collaborations. In the small countries, like Slovenia and the Slovak Republic, new institutions with new objectives were formed and some of the expertise has been diluted. The system of financing science only by projects has lead to a reduction of the facilities. There are even cases where within a big project new equipment has been purchased but due to the lack of further funding further use of the equipment has been impossible. Such demerits of the project funding system are also seen in Western European countries. The project funding system stimulates competitiveness, but when institutional funding is available in addition, the effect is better. Air Pollution Modelling As pointed in Zavodski et al (2000) a little effort has been made in all countries towards urban chemistry modelling. A photochemical module using the EMEP chemical reaction scheme was developed and tested in the Slovak Hydrometeorological Institute. At the Charles University in Prague a photochemical urban scale model SMOG was developed. A simple chemical scheme (21 reactions) was implemented in the model. The US EPA model OZIPR (Ozone Isopleth Plotting Package, R-research version), including EKMA resp. CBM schemes, was implemented in the Institute for Atmospheric Physics in Budapest. A combination of the causal model and a statistical model is being tested for forecasting of high ozone concentrations in Budapest. Several stochastic models, based on mutli-parametric regression, artificial neural networks, Kalman filter and other statistical techniques, were implemented for short-term forecast of air pollution episodes, namely high ozone concentrations (Czech Republic, Hungary, Poland, Slovenia). The new generation of 3D urban photochemical transport and dispersion models are not in use in East European countries mainly for two reasons, i.e. shortage of funds available for research programmes and the absence of appropriate emission and meteorological input data. The existing computational facilities were not sufficient and thus the advanced meteorological preprocessors were not usable. The situation is gradually changing due to the advances in computer development, which made them accessible. However, concerning the meteorological pre-

334

E. Batchvarova et al.

processing the situation is different. Limited area meteorological prognostic models were put into operation in some of the countries. French limited area hydrostatic models ALADIN (version Central Europe - LACE and version Southeast Europe - SELAM) are available at the meteorological services of Bulgaria, the Czech Republic, Hungary, Romania, Slovakia and Slovenia. The German Weather Service non-hydrostatic mesoscale meteorological model is implemented in Bulgaria and Poland. Recently the Atmosphere Research Programme was launched at Warsaw University in Poland, a common effort of the Interdisciplinary Centre for Computer and Mathematical Modelling (equipped with CRAY T3D) and the Institute of Geophysics. Some sophisticated research air pollution models of various scales are being put in operation. Generally, because of the limited support of science in the all countries in the last decade, efforts to develop dispersion models at a national level are diminishing. The tendency to implement the well-established foreign models was clearly identified. The intensive research in air pollution modelling during the last years in some of the 10 countries is due to the fact that it was concentrated at the Meteorological Services and where the structure and expertise was retained and further developed (Bulgaria, Romania). Another factor for further developments in the modelling studies is the existence of international structures like WMO and EMEP where all 10 countries participate. Participation in International Projects, Overview for all 10 Countries Framework Programme 4 of the EC The participation of the 10 countries in FP4 was under a separate budget and special conditions. The connections among the researchers was poorer than in the period of FP5. The 10 countries participated in only a few research projects. Framework Programme 5 of the EC Under the environment calls in the Energy, Environment and Sustainable Development section 267 projects were funded. About 150 of them were related with atmospheric research. Among those about 50 consortiums include one or more countries of the 10, including the accompanying measures projects such as conferences, centres of excellence and institutional projects. Most successful in the accompanying measures (8 projects) were the teams from Poland (Batchvarova et al., 2005). Bulgaria and the Czech Republic coordinated one project each. Considering the 35 multi-national research consortiums, German groups coordinated 12, English groups 7 and French groups 4 of them. Framework Programme 6 of the EC The database of FP6 is under development and therefore it is difficult to collect data on the participation of the 10 countries in atmospheric chemistry related projects. EC Twinning Programme All countries participated in twinning projects, but on different issues of the accession activities. Most active in such collaborations were the German institutions and groups.

Survey on Atmospheric Chemistry Research in Some New EU Member States

335

EUROTRAC-2 Almost all of the 10 countries participated in EUROTRAC-2 projects or subprojects. The EUROTRAC-2 Conferences were supporting participation of those countries, so presentations of groups not involved in specific EUROTRAC-2 projects were often given and thus the collaboration in the field of atmospheric chemistry enlarged. EMEP All 10 countries are members and participate in EMEP activities on Environmental quality measurements and modelling. WMO All 10 countries are members and participate in a number of programmes and activities on Environmental quality measurements and modelling, climate, weather forecast and others. The Meteorological services represent the countries in this collaboration. WHO All countries are members of the WHO and the Public Health services represent them. All participate in Environmental quality assessments, epidemiological studies and others. PHARE All countries were covered by the EC PHARE programme. The studies, projects and reviews were mainly orientated towards air quality monitoring networks. Conferences The research communities in all countries are very active in participating and organising international workshops, meeting, conferences, and congresses. Several Air Pollution Conferences were hosted in Bulgaria, Czech Republic, Poland, Hungary, and Slovenia. The Last European Aerosol Conference was held in Budapest, Hungary in 2004. A number of WMO and WHO meetings were organized by groups in these countries during the last 5-10 years. Synthesis of the Information and Identification of General Needs All countries have adopted national environmental strategies by the end of 2004, which are in line with the EU legislation and practice. In all countries the environmental legislation is in line with that of the European Union. In most of the countries the atmospheric chemistry process studies with appropriate facilities are limited. In the Czech Republic, Poland, Estonia and Slovenia such research is performed by only a few groups. In all the countries there are needs for greater research funding within the countries in order to be able to prevent the young scientists from emigrating. In all countries there is expertise and the senior scientists are aware of all the European and worldwide achievements in their fields. The modest national funding and existing facilities do not allow up-to-date atmospheric chemistry research, even such connected to monitoring. This brings the situation that individual researchers join European scientific projects and visit facilities. Rarely can these researchers provide comparable equipment which would allow them to participate within European projects on an equal basis. In addition, the national (in many cases modest) funding of research is now

336

E. Batchvarova et al.

organized on a project basis. This does not support the establishment of long-term modern facilities which are needed for contemporary research. The wide exchange of people and knowledge within Europe is extremely useful, stimulating and necessary. There is still more to do in order to bring researchers back to their country of origin. This need has been recognized in the last years with the new funding form Marie Curie or other fellowships granted to returning scientists. Still, there is much more to do within the countries, where the research groups are trying to keep the knowledge and to participate in international projects. Project funding has the disadvantage of short term, which makes difficult the support of expensive equipment. Even if some equipment is bought for a project, after the end of this project it is hardly used and there is no means to pay consumables and support personnel. The same problem exists in Western European countries as well. Although the aim of the survey was not to go into personal contacts with all groups, for some countries comments were requested from national experts (Slovenia, Hungary, Bulgaria, Czech Republic and Estonia). Also, discussions at conferences and workshops were used to form the conclusions and identify difficulties and gaps. Some of the identified difficulties, such as difficulties in communication among institutions within the countries, shortcomings of the project-funding research system and others, are observed in the Western European countries as well. Acknowledgements The survey was performed under contract 22382-2004-09 F1ED ISP BG between the EC JRC Ispra, Italy and NIMH, Sofia, Bulgaria. The authors, all from the Department of Composition of the Atmosphere and Hydrosphere at the National Institute of Meteorology and Hydrology, are thankful to Krasimira Lazarova from the same department for the technical help during the project. Colleagues from different countries were asked to comment on the survey of their country. The authors are grateful to them for their prompt response. References ANALYTIKA, 2004 - http://www.analytika.sk ARSO, 2005 - http://www.arso.gov.si/english/ Batchvarova E., T. Spassova, N. Valkov and L. Yordanova, 2004, Survey on atmospheric chemistry research in some new EU Member states and Candidate countries, Report on Contract 22382-2004-09 F1ED ISP BG between the EC JRC Ispra, Italy and NIMH, Sofia, Bulgaria. 99 p. Berlyand, M. (Ed.) 1981, Proceedings of the International Conference on Air Pollution held in 1977 in Leningrad, USSR, Leningrad, Gidrometeoizdat. BULAIR, 2002 - www.meteo.bg/BULAIR CHMI, 2004 - http://www.chmi.cz/uoco/isko/tab_roc/1997_enh/ENG/kap_02/kap_02_top.html CHTF, 2004 - http://www.chtf.stuba.sk/ EDXRS, 2002 http://iapf.physik.tu-berlin.de/EDXRS2002/sessions/postersessionii.htm

Survey on Atmospheric Chemistry Research in Some New EU Member States

337

EMEP, 2002 - http://www.emep.int/assessment/Part2/183-196_Part2-utan.pdf Estonian National Environmental Strategy (NES), Atmospheric Air, 2000 by Natalja Kohv, Valentina Laius, Siiri Liiv, Ott Roots, Reet Talkop Fiala, J. and J. Ostatnická (eds.), 2003, Air Pollution in the Czech Republic in 2002, Czech Hydrometeorological Institute, 158 p, ISBN 80-86690-07-5, www.chmi.cz/indexe.html. ICHF, 2005 - http://ichf.edu.pl/ IJS, 2005 – www.ijs.si Latvia, 1999, Report from Phare Country Visit- www.chmi.cz/uoco/isko/ptl/99ptl/qpr7-ann3.html Latvia, 2000, Report from Phare Country Visit - http://www.chmi.cz/uoco/isko/ptl/ qpr5ive.html Lithuania, 2001, Phare Country Visit Lithuania, Latvia and Estonia, 5-13 February 2001, Fifth Quarterly Progress Report, Annex IV.C, PTL/AQ - http://www.chmi.cz/uoco/isko/ptl/ Parvanova, M. 2003, Present conditions and problems of air pollution in Bulgaria, WHO Collaborating Centre for Air Quality Management and Air Pollution Control, Berlin, NL 30, http://www.umweltbundesamt.de/whocc/titel/ Puklová, V, P. Denková (eds.), 2003, Environmental Health Monitoring System in the Czech Republic – Report 2002 National Institute of Public Health, Prague 2003, 134 pages, ISBN 80-7071-215-5, www.szu.cz/chzpa/sumrep.htm REC, 1990, Regional Environmental Centre for Eastern and South-eastern Europe - www.rec.org/REC/ SHMI, 2004 - http://www.shmu.sk/ SHMI, 2004a - http://fsp.shmu.sk/document/MinutesBE12-16May.pdf Starnet, 2005 - http://starnet.rta.nato.int/results.asp?node=6&topic=38 SZU, 2004 - http://www.szu.cz/chzp/rep03/szu_03an/ka04_04.htm/ Twinning, 2000 - www.umweltdaten.de/luft/immissionen Ubelis, A. and V. Bergmanis, 2000, Insight in the history of point source air pollutant emission inventory in Latvia. In: Transport and Chemical Transformation in the Troposphere. Munich: EUROTRAC-2, 2000, TRAP-2. Zavodsky D, G Baranka, L Cernikovsky and Ken Stevenson, 2000, Review of urban air pollution models in PHARE accession countries for the support of CAFE October 2000 - http://www.chmi.cz/uoco/isko/ptl/

Appendix I. Addresses of Institutions Bulgaria Institution National Centre of Hygiene National Institute of Meteorology and Hydrology (NIMH), Bulgarian Academy of Sciences (BAS) Ministry of Environment and Waters (MEW) Geophysical Institute, BAS Faculty of Chemistry, Sofia University Faculty of Physics, Sofia University, Executive Agency on Environment, Ministry of

Address Blvd Dimitar Nestorov 15, 1431 Sofia, Bulgaria Blvd Tsarigradsko chaussee 66, 1784 Sofia, Bulgaria, http://www.meteo.bg Address: 1000 Sofia, 67 William Gladstone str., http://www2.moew.government.bg/ Acad. Bonchev Str., Block 3, 1113 Sofia, Bulgaria, www.geophys.bas.bg 1 Blvd James Baucher, 1126 Sofia, Bulgaria, www.chem.uni-sofia.bg 5 Blvd James Baucher, 1126 Sofia, Bulgaria, www.phys.uni-sofia.bg 136 Tzar Boris III blvd., P.O. Box 251

338 Environment and Waters University of Chemical Technology and Metallurgy (UCTM) University of Plovdiv “Paisii Hilendarski”, Faculty of Chemistry The Southwest University “Neofit Rilski” Institute of Physical Chemistry, Bulgarian Academy of Sciences (BAS) Institute of Catalysis, BAS Institute of General and Inorganic Chemistry, BAS

E. Batchvarova et al. 1618 Sofia, Bulgaria 8 St. Kliment Ohridski blvd..1756 Sofia, Bulgaria, www.uctm.edu 24 Tzar Assen St, 4000 Plovdiv, Bulgaria www.uni-plovdiv.bg 66 Ivan Mihailov str., 2700 Blagoevgrad, Bulgaria, www.swu.bg Acad. G. Bonchev Str., Block 11, Sofia 1113, http://www.ipc.bas.bg Acad. G. Bonchev St., Bldg.11, Sofia 1113, Bulgaria, http://ic.bas.bg Acad. G. Bonchev str. Block 11, 1113 Sofia Bulgaria, http://www.igic.bas.bg

Czech Republic Institution Charles University, Department of Meteorology and Environmental Protection The National Institute of Public Health, Department of Air Hygiene The Czech Hydrometeorological Institute (CHMI) Aerosol Laboratory, Institute of Chemical Process Fundamentals, Academy of Sciences of the Czech Republic J. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic The Czech Aerosol Society (CSA) Research centre of environmental chemistry and ecotoxicology (RECETOX) - Air Sampling Group (AIRSAG) Institute of Thermomechanics, Academy of Sciences of the Czech Republic

Address V. Holesovickach 2, 18000 Praque 8, Czech Republic, www.cuni.cz Srobarova st 48, 10042 Prague 10, Czech Republic, http://www.szu.cz/ Na Sabatce 17, 14306 Prague 4 – Komorany, Czech Republic, http://www.chmi.cz/indexe.html Rozvojova 2, CZ-165 02 Prague 6 – Suchdol, Czech Republic, http://www.icpf.cas.cz/aero/ Dolejškova 3, 182 23 Prague 8, Czech Republic, http://www.jh-inst.cas.cz Boþní II 1401, 141 31 Praha 4, Czech Republic, http://limail.ufa.cas.cz/en/info.html. Rozvojová 135, 165 02 Praha 6, Czech Republic, http://cas.icpf.cas.cz/index.php RECETOX - TOCOEN&Associates, Kamenice 126/3 625 00 Brno, Czech Republic, http://recetox.chemi.muni.cz/ Dolejkova 5, 182 00 Praha 8, Czech Republic, www.it.cas.cz

Estonia Institution National Institute of Chemical Physics and Biophysics Tartu University, Institute of Physical Chemistry Tartu University, Institute of Thechnology

Address 23 Akadeemia Rd., Tallinn 12618, Estonia, http://www.kbfi.ee/ 2 Jakobi St., Tartu 51014, Estonia http://www.ut.ee/index.aw/ Vanemuise 21, Tartu 51014, Estonia, http://www.tuit.ut.ee/

Hungary Institution The Ministry for Health and Social Welfare

Address Arany János u. 6-8, Budapest H-1051, Hungary, http://www.eum.hu/eum/eum_angol.main.page

Survey on Atmospheric Chemistry Research in Some New EU Member States The National Meteorological Service

Eötvös Loránd University – Dept. of Chemical Technology and Environmental Chemistry Eötvös University, Budapest, Department of Physical Chemistry Air Chemistry Group in the Hungarian Academy of Science, University Vezsprem Aerosol Association of Hungary

1024 Budapest, Kitaibel Pál u. 1. , Postal address:1525 Budapest, Pf. 38, Hungary, http://omsz.met.hu/english/kfo/l Pázmány P. s. 1/A, 1117 Budapest, Hungary, http://teo.elte.hu/fs/ chembase.html). Mail address: 1518 Budapest 112, P.O.Box 32. Hungary, http://www-phch.chem.elte.hu/, H-8201 Veszprém, P.O.Box 158, Hungary, http://www.mta.hu/index). 1085 Budapest, Rigo u. 3

Latvia Institution Ministry of Environmental Protection and Regional Development State Hydrometeorological Agency, Latvia Environmental Quality Division Latvian Environment Agency (under supervision of the Ministry of Environmental Protection and Regional Development) National Environmental Health Center (under supervision of Ministry of Welfare) University of Latvia, Institute of Atomic Physics and Spectroscopy Univ. of Latvia, Dept. of Analytical Chemistry

Address Peldu Str. 25, 1494 Riga, Latvia, www.gov.lv Tel: (371-7) 026-400, Fax: (371-7) 820-442 Maskavas Str. 165, 1019 Riga, Latvia Tel: (371-7) 144-390, Fax: (371-7) 145-154 Straumes str. 2, 2015 Jurmala, Latvia (371-7) 811-492, www.vdc.lv Klijanu Str. 7, 1012 Riga, Latvia Tel: (371-7) 379-231, Fax: (371-7) 339-006 19 Rainis Blvd., LV-1586 Riga, Latvia Tel./Fax.: +371 7229727; http://home.lanet.lv/~asi/ Riga, Latvia

Lithuania Institution Institute of Physics Ministry of Environment, Joint Research Centre, Environmental Quality Assessment Division

Address Savanoriu Pr. 231, 2053 Vilnius, Lithuania, www.fi.lt A. Juozapaviciaus 9, 2600 Vilnius, Lithuania

Poland Institution Institute of Physical Chemistry, Polish Academy of Sciences (PAS) Chemical Faculty, Technical University of Gdask Institute of Oceanology, PAS Environmental Sciences University of Mining and Metallurgy Voivodship Inspectorate of Environmental Protection - Slupsk Inst. of Occupat. Medicine and Environm. Health National Institute of Hygiene Institute for Ecology of Industrial Areas Wroclaw University of Technology Warsaw University of Technology Institute of Nuclear Chemistry and Technology

Address Kasprzaka 44/52, 01-224 Warsaw, Poland 80-952 Gdansk, Poland ul. Powstancow Warszawy 55, 81 712 Sopot, Poland Al. Mickiewicza 30, 30-059 Cracow, Poland Kniaziewicza 30, 76-200 Slupsk, Poland 13 Koscienla Street, 41-200 Sosnowiec, Poland 24 Chocimska, 00 791 Warszawa, Poland Ul. Kossutha 6, 40 833 Katowice, Poland ul. Smoluchowskiego 25, 50372 Wroclaw, Poland Nowowiejska 21/25, 00 005 Warszawa, Poland 16 ul. Dorodna, 03 195 Warszawa, Poland

339

340 Institute of Coal Chemistry, PAS The Inst. of Ecological Chemistry and Engineering Technical University of Gdansk

E. Batchvarova et al. Sowinskiego 5, 44 121 Gliwice, Poland Chlodnicza 4, 45 315 Opole, Poland G.Narutowicza Street 11-12 , P.O.Box 612 80 952 Gdansk, Poland



Romania Institution Institute of Hygiene, Public Health, Health Services and Management “Alexandru Ioan Cuza“ University of Iasi, Faculty of Chemistry University of Ploiesti PETROLEUM-GAS National Institute of Meteorology, Hydrology and Water Management National Institute for Marine Research and Development “Grigore Antipa”

Address Leonte str. 1-3, 76256 Bucharest, Romania Carol I Boulevard, 11, Iasi 6600, Romania. Bd. Bucuresti 39, 2000 Ploiesti, Romania Sos. Bucuresti-Ploiesti 97 71552 Bucuresti, Romania Mamaia Boulevard 300 8700 Constantza, Romania



Slovak Republic Institution The Hydrometeorological Institute (SHMU) Slovak Technical Univ., Dept. of environmental sciences, Faculty of Chemical and Food Technology Institute of Preventive and Clinical Medicine Department of Analytical Chemistry, Faculty of Natural Sciences, Comenius University

Address Jeséniova 17, 833 15 Bratislava http://www.shmu.sk/ Radlinského 9, 81238 Bratislava, Slovak Republic, http://www.chtf.stuba.sk/kei/indexe.php Limbová 14, 833 01 Bratislava, Slovak Republic, http://www.upkm.sk/, Mlynska Dolina 15, 1842 Bratislava, Slovakia, http://www.analytika.sk

Slovenia Institution University of Ljubljana, Faculty for Chemistry and Chemical Technology ECOSENSE Inst. of Public Health of the Republic of Slovenia Nova Gorica Polytechnic Josef Stefan Institute University of Ljubjana National Institute of Chemistry (NIC) Ministry of Environment and Spatial Planning, Environmental Agency of Slovenia AMES d.o.o.

Address Askerceva 5, SI-1000 Ljubljana, Slovenia Mirje 29, SI-1000 Ljubljana, Slovenia 2 Trubarjeva, 1001 Ljubljana, Slovenia Vipavska, 13 , PO Box 301, 5001 Nova Gorica, Slovenia 39 Jamova 39 , P.O. Box 3000 1001 Ljubljana, Slovenia Jamova 2, 1000 Ljubjana, Slovenia Hajdrihova 19, P.O.B. 660, SI-1001 Ljubljana, Slovenia Vojkova 1b, SI-1000, Ljubljana, Slovenia Na Lazih 30, 1351 Brezovica pri Ljubljani, Slovenia

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany: Input Data for Chemical Mass Balance Modelling Anita Niedojadlo, Karl Heinz Becker, Ralf Kurtenbach, and Peter Wiesen Department of Physical Chemistry, University of Wuppertal Gaußstr. 20, 42097 Wuppertal, Germany Key Words: NMVOCs, Oxygenated compounds, Emission profiles, CMB modelling, Source apportionment

Abstract Non-methane volatile organic compounds (NMVOC) were measured at various sites representing different areas and different emission sources in the city of Wuppertal, Germany. The measurements covered volatile hydrocarbons in the range of C2-C10 and oxygenated hydrocarbons such as alcohols, ketones and esters. Samples were collected using Carbotrap and Carbosieve SIII solid adsorption tubes and analysed off-line by thermal desorption and GC-FID analysis. Measurement results were used to create the input data for the source apportionment analysis with the Chemical Mass Balance Modelling technique. Emission profiles for traffic and solvent use were calculated. Introduction With respect to the European scale, road traffic and solvent use are by far the most important source categories of anthropogenic non-methane volatile organic compounds (NMVOC) emissions. Whereas NMVOC emissions from combustion processes contain predominantly pure hydrocarbons (alkanes, alkenes and aromatics), organic solvents and their vapours also contribute large quantities of oxygenated hydrocarbons such as alcohols, ketones, esters, glycol derivatives, ethers and halogenated hydrocarbons (Ullmann's Encyclopedia, 2001). Up to now, data from different German cities could only establish that road traffic represents the dominant source of shorter chain (C2-C9) hydrocarbons (Thijsse et al., 1999; Slemr et al., 2002; Gomes, 2002). On the other hand German NMVOC emission inventories show that solvent use is the main source of NMVOC emissions (Theloke et al., 2001; Umweltbundesamt, 2003). However, the relative importance of these two source categories is still afflicted with a large uncertainty mainly due to the lack of specific measurements of larger hydrocarbons and oxygenated species. In this work atmospheric concentrations of a large number of non-methane volatile organic compounds (NMVOCs) emitted by different anthropogenic sources, in particular from traffic exhaust and solvent use, have been investigated. The results from the studies should provide more information about the relative importance of road traffic and solvent use to the total NMVOC emission in Europe. Investigated compounds For better characterisation of both selected sources, namely traffic and solvent use, in addition to pure hydrocarbons also oxygenated species were included to the group of 341 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 341–350. © 2006 Springer. Printed in the Netherlands.

342

A. Niedojadlo et al.

investigated compounds. The source category of solvent use is characterised not only by a large variety of solvent-containing products but also by a broad range of applications in industrial production as well as in commercial and private use (Ullmann’s Encyclopedia, 2001). Among a large number of oxygenated species contained in solvents and solvent related products, the most common oxygenated compounds were targeted for the investigation. The choice was based on the known composition of products used in: water-based architectural coatings, consumer products, paint applications, printing industry, dry cleaning (McInnes, 1996; Friedrich and Obermeier, 1999; Ullmann’s Encyclopedia, 2001; EMEP/Corinair, 2003), automotive performance coatings (DuPont, 2001). The most abundant compounds were tested with respect to the capability of their qualification and quantification with the selected analytical method. The list of oxygenated compounds was additionally enhanced by methyl tert-butyl ether, a typical species added to gasoline as an octane enhancer in order to reduce emissions when gasoline is burned in engines (European Fuel Oxygenated Association, 2003). Finally, 18 oxygenated species were selected for investigation. Measurement procedure The measurements of non-oxygenated hydrocarbons and oxygenated species were performed according to the United States Environmental Protection Agency Compendium Method TO-17 entitled “Determination of Volatile Organic Compounds in Ambient Air Using Active Sampling Onto Sorbent Tubes” (Woolfenden and McClenny, 1999). The procedure was as follows: x x x x

ambient air collection by active sampling on glass tubes packed with adsorption materials, thermal desorption of sampled tubes, sample pre-concentration with a cryo trap, gas chromatography-flame ionisation detection analysis.

During the city air measurements sampling collection was performed for four air samples in parallel. Two samples were used for the analysis of non-oxygenated hydrocarbons and two for oxygenated hydrocarbons. For non-oxygenated hydrocarbons mulit-bed tubes packed with Carbotrap graphitized carbon and Carbosieve SIII carbon molecular sieve separated by glass wool were selected. For oxygenated hydrocarbons a combination of two adsorbent tubes with different characteristics was used, namely a tube packed with Carbotrap and a tube with Carbosieve SIII. Constructions of the adsorbent tubes are presented by Figure 1 and 2.

Figure 1. Construction of adsorbent tubes used for the sampling of non-oxygenated hydrocarbons.

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany

343

Figure 2. Construction of adsorbent tubes used for sampling of oxygenated hydrocarbons. The main characteristic of employed Carbotrap and Carbosieve SIII sorbents are presented in Table 1. Characteristics of the adsorption materials used for VOC sampling (Woolfenden and McClenny, 1999). hydrophobicity adsorbent max. temp (°C) specific area (m2/g) Carbotrap >400 100 yes Carbosieve SIII 400 800 no

Table 1.

After sampling the adsorption tubes were tightly capped and transported to the laboratory. Until the time of analysis the tubes were stored in a clean container placed in a refrigerator. Sampled tubes were analysed through a sequence of analytical steps. The analysis of non-oxygenated hydrocarbons and oxygenated species were performed separately. There were some variations in the manner of treatment of non-oxygenated and oxygenated compounds, however, the main steps were the same, namely a dry purge of the sampled adsorption tubes, thermal desorption of sampled tubes, analyte refocusing on a secondary trap and followed by GC-FID analysis. A schematic scheme of the analytical steps involved in the sample analysis is shown in Figure 3.

heated line

water trap

cryo trap

Preconcentrator 7100 Entech Instruments

capillary column

GC 6890 / GC 5840A Hewlett Packard

FID tube purge

tube desorption

focus

heated line

Thermal Desorption Unit 890 Supelco

outlet

carrier gas

Figure 3. Schematic overview of the sample analysis steps. The main difference in the way of analysis of non-oxygenated hydrocarbons and oxygenated species was the employment of 2 different gas chromatographs. The nonoxygenated hydrocarbons were analysed using a Hewlett Packard GC 6890 equipped with a 90 m HP-1 (100% dimethylpolysiloxane, nonpolar) capillary column. The oxygenated species were analysed with a Hewlett Packard GC 5840A using a 60 m DB-WAX (CARBOWAX,

A. Niedojadlo et al.

344

polar) capillary column. The temperature programs for both columns were optimised in order to obtain the best resolution. The FID’s parameters were also set to those giving the best sensitivity of detection. The parameters finally selected for the GC systems are presented in Table 2. Table 2.

Parameters of the GC-FID process.

Parameters capillary column x film material x column length x column diameter x film thickness carrier gas inlet temperature column pressure (setpoint) oven parameters during GC run: x start temperature x ramp x end temperature x total GC run x column back flashing detector (FID) parameters: x detector temperature x hydrogen flow x synthetic air flow x helium flow (make-up gas)

HP GC 6890 HP-1 (nonpolar) dimethylpolysiloxane 90 m (3 x 30 m) 0.32 mm 3.00 Pm helium 100°C 2 bar

HP GC 5840A DB-WAX (polar) carbowax 60 m 0.25 mm 0.25 Pm helium 100°C 2 bar

- 50°C over 10 min 5°C/min up to 200°C 200°C over 20 min 80 min on after 64 min

30°C over 30 min 5°C/min up to 200°C 200°C over 6 min 70 min -

300°C 40 ml/min 350 ml/min 30 ml/min

300°C 24 ml/min 340 ml/min 30 ml/min

The calibration procedure of non-oxygenated hydrocarbons was performed with a NPL (National Physical Laboratory) standard gas mixture containing 30 C2-C9 compounds. For oxygenated compounds, calibration was performed using pure liquid substances supplied by Aldrich, Lancaster or Merck. The calibration procedure involved: x substrate preparation in a 405 L reaction chamber under atmospheric conditions, i.e. 298 K and 760 Torr total pressure, x

determination of the substrate concentration by FTIR spectroscopy (Nicolet Magna 550),

x

active sampling onto adsorption tubes,

x

thermal desorption and GC-FID analysis.

This procedure allowed the calibration of the whole measurement system from the sampling tubes to the gas chromatograph. Measurement sites During three measurement campaigns performed in September 2001, August/September 2002 and October 2003, non-methane volatile organic compounds concentrations were measured at different city areas in Wuppertal near relevant sources of NMVOCs. The

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany

345

measurements were carried out using a car equipped with a hydrocarbon sampling system and automatic analysers for detection of carbon monoxide, carbon dioxide, nitrogen oxides, sulphur hexafluoride and a small met-station for meteorological parameters. In order to create the NMVOC emission profiles for road traffic, measurements were performed at areas representative for all major traffic conditions. Sampling was carried out in: x

a traffic tunnel (Kiesberg tunnel),

x

a down-town street intersection,

x

during drives in the city centre of Wuppertal and on the freeways.

It was assumed that the concentrations measured at the above sampling points and driving situations were dominated by emissions from traffic. For the NMVOC profiles of solvents emission measurements were performed around various solvent factories and workshops in Wuppertal. The following emitters were considered: x

DuPont Performance Coatings GmbH,

x

PPG Industries Lacke GmbH,

x

Bayer AG,

x

Dr. Alfred Conrads Lackfabrik Nachf. KG,

x

Karosseriebau Gorn GmbH.

To obtain the local ambient NMVOC concentrations sampling was performed at different points located in Wuppertal. The sampling sites represented residential, industrial, mixed settings and areas down-wind from the city centre. Chemical Mass Balance Modelling For the assessment of the contribution of the emission categories to the observed NMVOC concentrations the Chemical Mass Balance (CMB) modelling technique, version 8 from United States Environmental Protection Agency (Watson et al., 1998; Watson et al., 2001) was selected. The method uses source specific ratios between the emission rates of certain set of compounds and aims to recognise these fingerprints, or source profiles, in the profile measured at the receptor point. As a result the CMB model delivers contributions from each source type to the total ambient NMVOC and individual hydrocarbon species at receptor points and their uncertainties. The fundamental principle of the receptor models is that mass conservation can be assumed and the composition of source emissions are constant over the ambient and source sampling period. Therefore, the ratios between the components emitted by a single source are identical to the ratios between the resulting concentrations on the receptor location. Other assumptions required by the model are: x

the chemical species do not react with each other,

x

all sources which may significantly contribute to the receptor have been identified and their emissions characterised,

x

the number of source categories is less than or equal to the number of chemical species,

A. Niedojadlo et al.

346 x x

source profiles are linearly independent, i.e. sufficiently different from one another, measurement errors are random, uncorrelated and normally distributed.

These assumptions are fairly restrictive and difficult to fulfil in practice. However, the CMB model tolerates some deviations, which increase the final uncertainties of the source contribution estimation (Watson et al., 1998). The CMB calculations based on the input data presented below are the subject of a separate paper, which will be published elsewhere (Niedojadlo et al., 2005). Input data for the Wuppertal study As a model input to the CMB model serves: x

NMVOCs emission source profiles, so-called source fingerprints, which are the fractional amount of chemical species in the NMVOCs emissions from each source type,

x

receptor (ambient) NMVOCs concentrations, for which the contributions of emission categories are to be estimated,

x

realistic uncertainties for source and receptor values which are used to weight the relative importance of input data to model solutions and to estimate the uncertainty of the source contributions.

Because the purpose of this study was to provide more information about the relative importance of road traffic and solvent use to the total NMVOC emission only these two source categories were included in the CMB analysis. The fingerprints of NMVOC emission sources prepared for the CMB are those obtained from the source measurements. All fingerprints contain contribution from 105 species (95 varied concentrations, 9 pairs of co-eluted compounds). From about 190 hydrocarbons peaks detected in the GC-FID signals, the following compounds were selected for further investigation and CMB analysis: x

68 non-oxygenated hydrocarbons in the range C2-C10 from the hydrocarbon groups: alkanes, alkenes, alkynes, and aromatics,

x

18 oxygenated hydrocarbons in the range C1-C6 including alcohols, ketones and esters and also methyl tertiary butyl ether (MTBE),

x

19 hydrocarbon compounds with known carbon number but unidentified structure, these species were selected on the basis of their abundance and variation (compounds with an average concentration above 0.3 Pg/m3 and significant variation) and because of high significance to the source profiles diversification.

In this work the convention has been used that the fingerprints are normalised to unity. Results and discussion Traffic emission source profiles The measurements on the source profiles of traffic emission were carried out in a traffic tunnel (Kiesberg Tunnel), a downtown street intersection and during drives in the city centre of Wuppertal and on the freeways. Profiles of each of these measurements have been

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany

347

compared in order to establish whether they can be considered identical. Figure 4 presents the distribution of all the analysed compounds for all the traffic relevant measurements.

Figure 4. Emission profiles for road traffic; the mass distribution of 105 compounds. All the measured traffic profiles were found to be very similar. The same distribution could be observed for tunnel, intersection and street and freeway driving measurements. The highest contribution was from toluene, about 20%, with major contributions also from benzene, meta- and para-xylene, 2-methylpentane, isopentene, 1-butene and isobutene. Because of the good agreement between these profiles, they have been averaged to obtain a single traffic emission profile. This profile is considered characteristic of the traffic conditions and was used in the CMB analysis. The traffic fingerprint is presented on Figure 5.

Figure 5. Average emission profile of road traffic.

348

A. Niedojadlo et al.

Solvent use emission profiles To obtain solvent relevant emission profiles, measurements were performed around various solvent factories and workshops in Wuppertal. Due to the fact that the emission of solvent plants is not restricted only to the point emitters and is more like a surface emission, direct measurements of the emissions were not possible. Because of this, to determine the emission profile of particular factories the measurements of down-wind sources and background NMVOCs concentrations were performed. The solvent use fingerprints thus obtained are presented in Figure 6.

Figure 6. Emission profiles for solvent use; mass distribution of 105 compounds. As observed from a comparison of Figure 5 and 6, the solvent use fingerprints contain much higher contributions of oxygenated hydrocarbons than in the case for traffic. For example, for the measurements performed close to the DuPont factory (measurements on 18.09.01) the contribution of butyl acetate was about 80%. Also 2-butanol and acetone contribute significantly to the total mass. For non-oxygenated hydrocarbons xylenes and also toluene are important markers for the solvents emission. However, the solvent profiles obtained from the measurements differ significantly from one another. For different measurements different mass distributions were found, thus, no average solvent emission profile could be constructed. Receptors To perform the source apportionment the Chemical Mass Balance model applies the emission profiles from traffic and solvent use to the NMVOCs concentrations measured at the different receptors point in Wuppertal. Measurements were performed down-wind from the city centre, in areas close to the factories and workshop producing or using solvents, in dense traffic areas and in residential areas. The characteristics of the receptor points are presented in Table 2.

New Measurements of NMVOC Concentrations in the City Air of Wuppertal, Germany Table 2.

349

Receptor points for CMB analysis. 6 NMVOC measured

receptor sites

characteristic

sampling time

Blücher Str.

Hatzfelder Str.

Industrial area, close to PPG solvents factory, down-wind from the object Industrial area, close to DuPont solvents factory Down-wind from the city centre of Wuppertal Down-wind from the city centre of Wuppertal City centre of Wuppertal, dense traffic intersection City centre of Wuppertal, dense traffic intersection Residential area outside from the city centre Residential area outside from the city centre Industrial area, close to DuPont solvents factory Industrial area, close to DuPont solvents factory

Simon Str.

Bissing Str.

(Pg/m3)

r

18.09.01 13:20-14:20

53.809

2.694

18.09.01 15:20-16:20

515.607

7.590

19.09.01 11:10-12:10

29.775

2.629

22.08.02 15:06-16:06

31.083

2.422

26.08.02 15:35-16:35

58.344

2.167

26.08.02 16:40-17:40

42.527

3.235

27.08.02 15:55-17:05

30.610

1.524

27.08.02 17:12-18:24

23.827

1.190

29.08.02 13:06-14:10

18.408

3.061

29.08.02 14:55-15:55

45.064

2.842

Industrial area, close to Bayer factory

03.09.02 10:35-11:40

26.104

3.071

Industrial area, close to PPG solvents factory Industrial area, close to PPG solvents factory Industrial area, close to Conrads solvents factory Industrial area, close to PPG solvents factory, up-wind from the object Industrial area, close to PPG solvents factory, down-wind from the object

03.09.02 12:52-13:52

36.890

4.112

04.09.02 14:29-15:30

23.233

1.398

04.09.02 16:17-17:17

46.234

2.377

13.10.03 13:06-14:06

21.895

1.930

13.10.03 13:06-14:06

27.548

2.145

Märkische Str.

Close to free way 46

15.10.03 13:08-14:08

72.789

8.992

Hatzfelder Str.

Industrial area, close to DuPont solvents factory Industrial area, close to Gorn solvents workshop Industrial area, close to Gorn solvents workshop

15.10.03 12:55-14:00

61.431

5.503

17.10.03 09:45-10:45

115.315

11.772

17.10.03 09:45-10:55

75.803

10.048

Hatzfelder Str.

Girardet Str. Uni

Bundesallee Bundesallee Im Johannistal, Im Johannistal, Wilkhaus Str.

Lützow Str.

Viehhof Str. Yorck Str. Bissing Str.

Simon Str. Simon Str.

Conclusions NMVOC profiles for traffic emissions were measured in a traffic tunnel, a down-town intersection, during city and freeway drives. All the profiles obtained showed very similar mass distributions of the measured compounds. Hence, one single averaged traffic emission profile representing all traffic conditions was created. Solvent emission profiles were measured in the vicinity of several solvent factories and workshops in Wuppertal. Contrary to the emissions from traffic, the solvent fingerprints obtained from different measurements indicate significant differences between the emissions of different solvent relevant sources. According, it was not possible to create one average solvent use source profile and particular solvent emitters were considered separately. Significant differences between the major compound concentrations for traffic and solvent use emission profiles could be observed. Compared to the traffic fingerprints the

350

A. Niedojadlo et al.

emission profiles representing solvent use had much higher contributions of oxygenated compounds and toluene and xylenes. Further measurements and calculations are necessary in order to improve the emission profiles of solvent use and to include additional emission sources such as evaporative losses of motor fuel, natural gas leakage, stationary fuel combustion and biogenic emissions to the CMB apportion analysis. Acknowledgements The financial support by the Deutsche Bundesstiftung Umwelt (German Environment Foundation - DBU) is gratefully acknowledged. References DuPont Performance Coatings GmbH, personal communication (2001). European Environmental Agency; Emission inventory guidebook, 3rd Edition, Copenhagen (2003). European Fuel Oxygenates Association; MTBE resource guide (2003). Friedrich, R., A. Obermeier; Antropogenic emissions of volatile organic compounds, in: C. N. Hewitt (eds), Reactive hydrocarbons in the atmosphere. Academic Press, San Diego (1999) 2-38. Gomes J.A.G.; Impact of road traffic emissions on the ozone formation in Germany, PhD thesis of University of Wuppertal, Wuppertal (2002). McInnes, G. (eds); Joint EMEP/CORINAIR atmospheric emission inventory guidebook, European Environmental Agency, Copenhagen (1996). Niedojadlo, A., K.H. Becker, R. Kurtenbach, P. Wiesen; New measurements of NMVOC concentrations in the city air of Wuppertal: Source apportionment by Chemical Mass Balance Modelling, Atmos. Environ. manuscript in preparation (2005). Slemr, F., G. Baumbach, P. Blank, U. Corsmeier, F. Feidler, R. Friedrich, M. Habram, N. Kalthoff, D. Klemp, J. Kühlwein, K. Mannschreck, M. Möllmann-Coers, K. Nester, H.J. Panitz, P. Rabl, J. Slemr, U. Vogt, B. Wickert; Evaluation of modelled spatially and temporarily highly resolved emission inventories of photosmog precursors for the city of Augsburg: The experiment EVA and its major results, J. Atmos. Chem. 42 (2002) 207-233. Theloke, J., A. Obermaier, R. Friedrich; Abschätzung der Emissionen von Lösemitteln in Deutschland, Gefahrstoffe Reinhaltung der Luft 61 (2001) 105-112. Thijsse, Th.R., R.F. Van Oss, P. Lenschow; Determination of source contributions to ambient volatile organic compound concentrations in Berlin, J. Air and Waste Manage. Assoc. 49 (1999) 1394-1404. Ullmann's Encyclopedia of Industrial Chemistry, Sixth Edition, Electronic Release (2001). Umweltbundesamt; Umweltdaten Deutschland Online (2003). Watson, J.G., J.C. Chow, E.M. Fujita; Review of volatile organic compound source apportionment by chemical mass balance, Atmos. Environ. 35 (2001) 1567-1584. Watson, J.G., N.F. Robinson, E.M. Fujita, J.C. Chow, T.G. Pace, Ch. Lewis, Th. Coulter; CMB8 applications and validation protocol for PM2.5 and VOCs, Desert Research Institute Document No. 1808.2D1, U.S. Environmental Protection Agency (1998). Woolfenden E.A., W.A. McClenny; Determination of volatile organic compounds in ambient air using active sampling onto sorbent tubes, Compendium Method TO-17, Center of Environmental Research Information, U.S. Environmental Protection Agency EPA/625/R-96/010b, Cincinnati (1999).

Surface and Total Ozone Over Bulgaria Staytcho Kolev and Vera Grigorieva National Institute of Meteorology and Hydrology, Bulgarian Academy of Sciences, 66 Tzarigradsko Chaussee, 1784 Sofia, Bulgaria Key Words: Stratospheric intrusion, Solar eclipse, Photochemical O3 formation

Introduction The ozone problems (stratospheric, tropospheric and surface O3) are an important part of atmospheric air pollution problems. One of the goals, outlined in the Recommendations of the meetings of the Ozone Research Managers (under World Meteorological Organization – WMO) was: the conducting of systematic measurements, which provide the basis for understanding the ozone regime and its trends. Total ozone In Bulgaria, the first total ozone measurements with a Dobson ozone spectrophotometer started only in the early 1960s in Sofia, and were carried on for about 5 years. Then, after a few years break, the Bulgarian National Institute of Meteorology and Hydrology (NIMH) in Sofia started to use Russian filter ozonometers. In 1998, two Russian ozonometers M-124 were renovated and calibrated at the Main Geophysical Observatory in St.Petersburg, with financial support from the WMO. All collected data are sent every month to the WMO World Ozone and UV Data Center in Toronto. Figure 1 shows the typical monthly variation of total ozone over Sofia, and compares the O3 behaviour over Sofia to that over Potsdam and over Rome. Since the beginning of the last decade, total ozone observations have also been carried out in Stara Zagora town. Vertical ozone sounding In the period 1983-1992, balloon ozone soundings were released weekly at the NIMHSofia. For that purpose were used ozonesondes, manufactured in Eastern Germany. The activities were interrupted largely due to financial difficulties resulting from the transition to a market economy. Tropospheric ozone is a major atmospheric pollutant, which plays a key role in atmospheric chemistry, and at heightened levels causes damage to human health, forest ecosystems, agricultural crops and materials. In Europe, the systematic long-term observations of surface ozone have been carried out at latitudes higher that 45o, while comparatively less is known about the ozone behaviour near the ground in Southern and South-Eastern Europe (Lindskog et al., 2003). Initial investigations of the surface ozone in Sofia, Bulgaria, started in 1994. The following main goal was pursued – to study the surface ozone behaviour and the processes that define this behaviour. This paper presents the most interesting results on the surface ozone behaviour in Sofia.

351 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 351–358. © 2006 Springer. Printed in the Netherlands.

S. Kolev and V. Grigorieva

352

Figure 1.

Monthly mean values of total ozone over Sofia in 2000 and 2001 (top), and the comparison of monthly mean values of total ozone over Rome, Potsdam and Sofia (bottom).

Site and instrument description The measurements were performed in Sofia (42q39c N, 23q23c E, ca 588 m a.s.l.), which is located in the western part of Bulgaria. The observation site is situated about 7 km to the southeast of the city centre, and has a fairly good vegetative ground cover. At 100 m away from the site there is a road with considerable car traffic. The ozone recorder was installed about 10 m above the ground level. A Teflon pipe was used for air sampling. The measurements were performed mostly during daylight hours, and less regularly in twenty-four hours period.

Surface and Total Ozone Over Bulgaria

353

The ozone detector used in the investigations was a chemiluminescent ozone analyzer, model 3-02P1, by OPTEC Inc. (Russia). The measuring principle was that the sensor, consisting of an organic dye on the solid-state support, developed chemiluminescence in the presence of ozone. It must be noted that the chemiluminescence ozone measurements are considerably more free from interferences from other atmospheric components than are the ultra-violet (UV) optical absorption measurements. The UV optical absorption technique of surface ozone detection, which was tested with clean calibration mixtures under laboratory conditions, can give a positive offset in polluted air measurements (Grigorieva and Mihalev, 2003a). The ozone analyzer used in our observations had the response time of not more than 1 s, and the sensitivity of about 2 Pg/m3. Periodically, the analyzer was calibrated against an external O3 generator. Results and discussion Ozone variations and meteorological conditions The pattern of diurnal variations of the surface ozone concentrations is strongly influenced by meteorological conditions. The pronounced O3 maximum in the daytime, which is explained in terms of vertical mixing process and photochemical ozone production, occurred on clear windless afternoons. During the fine windy weather the dilution of the atmospheric pollutants takes place. So the decreased ozone concentrations are detected and the ozone level is approximately constant throughout the day. However, in the cases when vertical exchange is limited (autumn-winter period, nocturnal inversions) the wind enhances the vertical mixing (longrange transport of the polluted air mass also can takes place) and increases the ozone content near the ground. The cloudiness strongly decreases the ozone concentrations near the ground, but when it is foggy the ozone content is very low, often zero. So the ozone concentrations sensitively reflect the meteorological conditions under which measurements are performed. It is very like that more realistic information about temporal and spatial ozone variations may be obtained if ozone data recorded under similar meteorological situations are analyzed. Ozone pollution in Sofia The annual surface ozone behaviour clearly shows a seasonal variation with a summer maximum, which usually varies between ca 60 and 100 µg/m3, depending on the meteorological circumstances. The minimal ozone concentrations (20-35 Pg/m3) were detected during the winter period. The episodes with high O3 concentrations, up to 140 µg/m3, are sometimes observed in summer periods. The recorded seasonal pattern of ozone concentrations is similar to that observed at northern mid-latitudes. The collected data show, that the surface ozone concentrations in Sofia do not exceed the European Union standard hourly average of 180 Pg/m3. It is interesting to compare these results to ozone pollution in some other cities of the Balkan Peninsula, for which data are available. For example, in a big Romanian city Timisoara the maximum ozone concentrations during summer months of 1994-1996 reached only 116-122 µg/m3 (Lorinczi et al., 1999). In western Croatia, the summer ozone pollution in 19961997 rarely exceeded 120 µg/m3 (Butkovic et al., 1999). In Thessalonica (northern Greece), the maximum 8-hours concentration of O3 in 1995 was 90 µg/m3 – the lowest value among a number of large European cities (EEA, 1999). The results are consistent with model calculations, which show that although the efficiency of the photochemical ozone production (the number O3 molecules per NOx molecule) is higher in southern Europe than in western Europe, the chemical

S. Kolev and V. Grigorieva

354

ozone formation per unit area is more intensive in the western part of the continent, due to the high concentrations of ozone precursors (Roemer et al., 1997). In a general case, it is very difficult to evaluate the role of individual processes, such as long-range transport of the polluted air, local photochemical ozone formation and direct stratospheric intrusions of ozone rich air, in episodes with high ozone pollution. Thus, each particular case was analysed in detail, to identify possible sources of the heightened ozone pollution in Sofia. Long-range transport Ozone pollution on June 3rd 1994, with maximum hourly O3 concentration up to 140 µg/m , was investigated (Grigorieva and Mihalev, 2003b). The event was studied taking into account the ozone data base over Europe during the last week of May 1994. The highest ozone concentrations in Europe in 1994 – 490 µg/m3 (hourly average) – were measured in Italy on May 24th (EEA, 1996). The analysis of air flow at our site showed that in the end of May the prevailing wind direction was from the west sector, i.e. from Italy, (Figure 3b). Just on June 3rd, the meteorological conditions (sunny day, windless until 17:00 LT) were favourable for observation of the variations of ozone concentration, which had a distinct maximum (Figure 2b), was associated with the well-mixed layer and contained information about the ozone concentration in the upper part of the boundary layer, in which the long-range transport occurs. For comparison, Figure 2a shows the behaviour of surface ozone on May 17th, a day with similar meteorological conditions but different wind directions to preceding days (Figure 3a). So there are reasons to assume that the episodes with elevated ozone concentrations in Sofia may be associated with the long-range transport of polluted air masses form the west and the south-west sectors. It should be noted that the highest ozone concentrations observed in Croatia, Slovenia and Hungary are also associated with western, south-western and north-western winds (Hov, 1997). 3

Stratospheric intrusions Initial investigation of the possible role of the direct stratospheric intrusions of the ozonerich air during the high O3 episodes observed near the ground was performed (Grigorieva et al., in press). The origins of high ozone concentrations (year extremes) registered in Sofia during summer 1998 were analyzed, taking into account the information on stratospheric intrusions. The O3 content in the atmospheric air reached values of up to 125 µg/m3 on July 27th, and up to 157 µg/m3 on August 5th. It must be noted, that there was a cold wind on first ten days of August, which normally are warm. Analysis of the data showed that during the same periods, the year extreme ozone concentrations were registered in a large part of Europe. Stations located close to Sofia registeres the following values: K-Puszta, Hungary – the year maximum of 255 µg/m3 on August 4th; Krvavec, Slovenia – high values up to 187 µg/m3 on July 27th-31st, and up to 150 µg/m3 on August 3rd-4th; Kovk, Slovenia – a maximum of 138 µg/m3 on August 3rd. The Mt. Cimone station in Italy showed maximums on July 24th (232 µg/m3) and on August 10th (250 µg/m3). Most of the mentioned values were the absolute year maximums. Taking into account the information about the intensity of stratospheric intrusions during this period (Bonasoni, 2001), one can assume that the large scale stratospheric intrusion events, several days long and covering a vast territorial range, took place over Europe during the considered period.

Surface and Total Ozone Over Bulgaria

355

a

Figure 2.

Daytime variation of the surface ozone concentration, wind speed and temperature on May 17th 1994 (a) and June 3rd 1994 (b).

a

Figure 3.

b

b

The wind roses in the days before May 17th 1994 (a) and June 3rd 1994 (b). The lengths of the cuts are proportional to the frequency of the wind occurrence in the given direction.

S. Kolev and V. Grigorieva

356 Photochemical ozone formation in-situ

The rapid and regular changes of sunlight radiation during a solar eclipse are a rare opportunity to study under real atmospheric conditions the specific behaviour of ozone and, in particular, the ozone production in fast photochemicl reactions proceeding in the boundary layer. The strong influence of meteorological conditions, not only cloudiness but also wind, on the O3 variation near the ground appears to explain why only a few well time-resolved observations of the surface ozone variation during solar eclipse are available up to now (see Grigorieva and Gogosheva, 2003). The meteorological situation that took place in Sofia, Bulgaria during the 94% solar eclipse on 11th August 1999 occurs extremely rarely: x practically clear sky over the day, x practically no wind over the day, x time of the solar eclipse coincided with the main period of photochemical ozone formation.

Figure 4.

Behaviour of the surface ozone concentration during the 94% solar eclipse in Sofia on August 11th 1999 (curve B) and on August 18th 1999, a control day with close meteorological conditions (curve A).

Figure 4 shows the surface ozone behaviour in Sofia during solar eclipse (curve B) and compares it to that observed on 18th August 1999, a control day with close meteorological conditions (curve A). It is clear that O3 distribution is strongly affected by the solar eclipse. The curves representing the variation of surface ozone concentration and the solar radiation during the eclipse are very similar in shape (Figures 4 and 5, respectively). A short time after

Surface and Total Ozone Over Bulgaria

357

first eclipse contact, the surface ozone concentration still does not "feel" the solar eclipse. The further fall of the sunlight intensity decreases the efficiency of the photochemical ozone formation in-situ, so fast diminution of ozone content in the atmosphere near the ground is observed. The maximum decrease of O3 concentration – from ca 90 µg/m3 to ca 55 µg/m3 – was detected approximately 10 minutes after the totality. This lag is related to the loss rate of ozone (including O3 destruction by NO) and to the photochemical production rate after the maximum phase. The unique high-quality recording of surface ozone variations during a solar eclipse allows the visualization of the dynamics of ozone production and destruction in the boundary layer, and an estimatation the amount of ozone formed due to fast in-situ photochemistry. The opportunity to observe both processes under real atmospheric conditions is extremely rare.

Figure 5.

Variations in UV solar radiation observed in Stara Zagora (about 200 km away from Sofia) during solar eclipse on August 11th 1999.

Conclusions x In the polluted atmosphere, the ultra-violet optical absorption technique (UV-photometer) does not always provides quality ozone data. A possible influence of various interferences on the measuring procedure can be responsible for this. The chemiluminescence method of O3 detection is considerably freer from interferences. x The unique, highly time-resolved variations of the surface ozone concentration during a solar eclipse were recorded, which demonstrate the occurrence of fast photochemical reactions in the boundary layer. The efficiency and the rate of the photochemical ozone production in-situ

S. Kolev and V. Grigorieva

358

under real atmospheric conditions was clearly demonstrated. These observations show that at our site during summer (under ordinary pollution and meteorological conditions, that are favourable for photochemical ozone production) ca 40% of the measured ozone is produced photochemically in-situ. The O3 concentrations which exceed ca 100 - 110 µg/m3 can be attributed to long-range transport of polluted air masses or to stratospheric intrusions. x Stratospheric intrusions of ozone-rich air can be registered not only by high-altitude mountain stations, but also by stations at low elevations (including the sub-urban locations). x During the analysis of summer episodes of high ozone concentrations, it must be kept in mind that they can have not only photochemical origin, as it is often assumed, but also stratospheric origin. Acknowledgements This work was supported by the Bulgarian Ministry of Education and Science, under contract number NZ-1406. References Bonasoni, P.; Study of tropospheric ozone at Mt. Cimone, in: A. Lindskog (ed), Tropospheric ozone research. Annual Report 1999, EUROTRAC-2 ISS, Munich (2001) 54-58. Butkovic, V., T. Cvitas, N. Kezele, L. Klasinc and S. Vidic; Boundary layer ozone levels in the Adriatic region, in: P.M. Borrell, P.Borrell (eds), Proc. EUROTRAC Symposium’98, WIT Press, Southampton (1999) 281-285. EEA (European Environment Agency); Air quality in Europe, 1993 – a pilot report, Topic Report No 25. EEA, Copenhagen (1996). EEA (European Environment Agency); Environment in the European Union at the turn of the century, Environment Assessment Report No 2. EEA, Copenhagen (1999) 139. Grigorieva, V. and M. Mihalev; The possible uncertainties over quality of the surface ozone data, Proc. SPIE Int. Soc. Opt. Eng. 5226 (2003a) 285-289. Grigorieva, V. and M. Mihalev; Studies of surface ozone over Sofia, Bulgaria, Izvestiya RAN, Atmospheric and Oceanic Physics, 39, Suppl.1 (2003b) S51-S55. Grigorieva, V. and T.Gogosheva; Variations in surface ozone concentration during the solar eclipse August 11, 1999, Izvestiya RAN, Atmospheric and Oceanic Physics, 39, Suppl.1 (2003) S47-S50. Grigorieva, V., S.Kolev and M.Mihalev; Investigation of correlations between the high surface ozone episodes and the stratospheric intrusion events, Proc. SPIE Int. Soc. Opt. Eng. (in press). Hov, O. (ed); Tropospheric ozone research, Springer, Heidelberg (1997) 222, 271, 317. Lindskog, A., M. Beekmann, P. Monks, M. Roemer, E. Schuepbach and S. Solberg; Tropospheric ozone reseaech, in: P. Midgley, M.Reuther (eds), Towards cleaner air for Europe – science, tools and applications. Part 2. Overviews from the final reports of the EUROTRAC-2 subprojects. Margraf Verlag, Weikersheim: (2003) 251-270. Lorinczi, E.A., M. Lorinczi and E. Peter; Monitoring surface ozone at Timisoara, Romania, in: P.M. Borrell, P. Borrell (eds), Proc. EUROTRAC Symposium'98, WIT Press, Southampton (1999) 911-914. Roemer, M.G.M., R. Bosman, T. Thijsse, P.J.H. Builtjes, J.P. Beck, M. Vosbeek and P. Esser; Budget of ozone and precursors over Europe, in: O. Hov (ed), Tropospheric ozone research. Springer, Heidelberg (1997) 461-467.

Heavy Metals Pollution: An Everlasting Problem Raluca Mocanu, Simona Cucu-Man, and Eiliv Steinnes Department of Analytical Chemistry, Faculty of Chemistry, ‘Al. I. Cuza’ University, Iasi, Blvd. Carol I, no. 11, 700506 Romania Key Words: Bio-accumulation, Biomonitoring, Heavy metals, Moss types

Introduction The “heavy metal in the environment” collocation refers to any metallic chemical element and some metalloids (e.g. arsenic) that are toxic or poisonous for living organisms even at low concentration, e.g. Pb, Cd, Hg, As, Tl, Cr. They originate in the Earth’s crust as well as in the majority of wastes resulting from anthropogenic activities. Toxic effects of other heavy metals (Cr, Mo, Ni, As, Se etc.) have to be considered separately from the effects of biologic doses in which they exert their vital role. The danger represented by heavy metals lies in three aspects of their behaviour once the reach organisms: 1. they cannot be degraded or transformed into harmless products; 2. they cannot be destroyed and 3. they are bio-accumulative leading to an increase of their own concentration in a living body (animal, plant, human) over time. Thus, accumulated compounds in living organisms are stored faster than they are metabolised and excreted. The four most dangerous metals and metalloids, not occurring naturally and having no known function in biochemistry and physiology of living organisms are: Pb, Cd, As, Hg. Several examples of the health risks due to the penetration of these heavy metals into a living organism are listed below: Effects of Cadmium (Cd): its chemical similarity to zinc (an essential micronutrient for animals, plants and humans) leads to its toxicological properties. Cd once absorbed by an organism remains stocked for many decades. This bio-persistence generates lung diseases, bone defects, renal dysfunction, increases blood pressure etc. Effects of Lead (Pb): it is one out of four metals that have the most dangerous effects on living organisms. It can enter the human body via air, water and in most of all through uptake via food. In humans Pb can be associated with kidney and brain damages, a rise in blood pressure, disruptive child behaviour and learning abilities etc. Effects of Mercury (Hg): it is toxic by inhalation, ingestion and skin absorption. The effects of long-term exposure to Hg include: increase of blood pressure, hallucinations, memory loss etc. Hg does not have a known function in living organisms. th

Effects of Arsenic (As): known as far back as the 13 century, As is a poison as well as a cure for some diseases or metal for glass, laser, semiconductors. Routes of attack are ingestion, inhalation, skin and eye contact. The targeted organs are: digestive, respiratory, reproductive and central nervous systems, skin, liver, kidneys etc. 359 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 359–368. © 2006 Springer. Printed in the Netherlands.

R. Mocanu et al.

360

Besides the above there are some elements whose presence, if the concentration exceeds the permitted limits, can be considered as pollutants and can become toxic. Effects of Copper (Cu): known to the ancients and most often used as an electrical conductor, in art, industry, and coins. Cu enters the human body via skin or/and eye contact, inhalation, and ingestion. It causes damage to the following target organs: skin, eyes, liver, kidneys and the respiratory system. Copper deficiency causes anaemia, growth inhibition etc. Effects of Zinc (Zn): it plays a remarkable role in the metabolic cycles of human and animals and it is an essential element for plant growth. No toxic effects of Zn are known. It helps copper uptake in human and animal organisms. In addition to the above-mentioned metals, essential or not for human life, mention should also be made of toxic metals like chromium, silver, selenium, barium, aluminium (if their concentration goes beyond the permissible exposure limit). Heavy metals are usually available in the structure of the Lithosphere. Together with those originating from human activities they can enter living organisms through water, air and the food chain (Figure 1). Atmospheric input Continental air Oceanic air

Litosphere and ocean

Aquatic input Running waters Stagnant waters Sediments (sink) (streams, rivers) (lakes, swamps, pools)

Figure 1.

Heavy metals’ cycle in the environment.

A UN-ECE Protocol of heavy metal atmospheric emissions control was signed in 1998. This document designated cadmium (Cd), mercury (Hg) and lead (Pb) as first priority toxic elements. In point of fact the Protocol, in contrast to others as far as the water, soil or sediment pollution with heavy metal is concerned, is the unique “screen” of atmosphere protection. Legislation does not yet exist for other elements with respect to defining the atmospheric pollution by heavy metals. It is left to each country to respect and to protect the atmosphere from pollution through heavy metals.

Heavy Metals Pollution: An Everlasting Problem

361

Prut River’s basin (Romania). A possible source of trans-boundary pollution with heavy metals Prut River is a first-order tributary of the Danube River; it is the last major tributary before the Danube enters its delta. The total length of the Prut River is 983 km. The uppermost 241 km are on Ukrainian territory, for the next 31 km the river forms the border between Romania and Ukraine, and for the lower 711 km it forms the border between Romania and the Republic of Moldova. The Prut River catchment is bordered in the South and West by the Siret catchment, situated on Romanian territory, and in the North and East by the Nistru catchment, situated on Ukrainian and Moldovan territory. Temporal and spatial information on trace metal (Cd, Cu, Pb and Zn) variation in different environmental compartments of the catchment were obtained by analysing samples collected from nine sites along the Romanian bank line (Figure 2b) during three years (20002002). 1. Description of site 1. Physical-geographical conditions: combination of steppe and forest-steppe. The climate is temperate continental: in winter the mean monthly air temperature is about 10 degrees below zero, in summer it is about 20 degrees above zero. Snow cover is extensive and reaches 30 – 50 cm. The relief is partially plane, partially covered with hills. The highest points are about 800 - 900 m above the sea level. The average duration of positive temperature period comprises 200 - 250 days a year. 2. Soils: the soil layer is represented by chernozem and chestnut soils with humus contents. Vegetation: agricultural crops, vineyards, orchards, meadows. 3. Hydrology: water system includes the Prut River with tributaries as well as freshwater lakes. 4. Industry: several cities, the main being Iasi.

a Figure 2.

b Monitoring sites along the Prut River.

R. Mocanu et al.

362 2. Water and sediment pollution with metals

Three very well known forms of pollution can affect water, one of the main sources of life: physical, chemical and biological pollution. The main sources of pollution with heavy metals are: 1. Natural sources –minerals, rocks. 2. Agricultural sources: pesticides, fertilizers that contain heavy metals. 3. Industrial sources: metal ore processing industries, leather, paint, textiles and paper factories etc. 4. Domestic wastewater. 5. Mine wastewater. Heavy metal contamination belongs to the chemical pollution category of persistent (slow or non-degradable) substances. The most dangerous chemical pollution is represented by waste chemical and agricultural drainage systems. Data from 2001 were selected so as to present typical concentration levels of the four metals along the Prut River (Figure 3). For Cu the variation along the river is rather small, but there is a tendency to higher values at sites 6 and 7 than at the other sites. In the case of Zn the variability of its concentration is greater between sites. Levels increase markedly at site 6 relative to the above sites. This is most probably due to input from one of the river tributaries, which receives the wastewater discharge from Iasi, the largest city within the Prut River catchment. Pb shows only moderate variation along the river course. There is a tendency to lower values at sites 8 and 9 compared to the upstream sites. This trend, which is not observed to a similar extent for the other metals, is probably related to a higher removal of Pb from the water column by sedimentation than for the other elements. For sediments, except for a few single samples, the distribution of metals is rather uniform along the river. The tendency to higher values at site 6 and below as observed for the water data is not clearly evident in the sediments (Figure 4). These trends indicate that the Prut River sediment generally is not heavily polluted with the four heavy metals.

Concentration of heavy metals in water Cu samples

Zn Pb

Concentration (P g/L)

25 20 15 10 5 0 0

1

2

3

4

5

6

7

8

9

Site

Figure 3.

-1

Annual mean concentrations of elements in the Prut River in 2001 (µg·L ).

Heavy Metals Pollution: An Everlasting Problem

363

Concentrations of heavy metals in sediment samples Concentration (mg/kg)

300

Concentration

Cd concentration in sediment samples

Cu Pb Zn

250

0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0

1

2

3

4 5 Site

6

7

8

9

200 150 100 50 0 0

Figure 4.

1

2

3

4

Site

5

6

7

8

9

-1

Metal concentrations in Prut River sediments collected in 2001 (mg·kg ).

3. Atmospheric pollution with heavy metals and their biomonitoring Surveys of air quality are progressively extending to areas at substantial distances from industrial and urban centres. In these areas the air pollution, frequently qualified as background pollution, is generally connected with the dispersion of pollutants on a large scale of time and space. Among the pollutants of interest, heavy metals are a special category considering the risks that they represent for human health and the danger caused by their accumulation in ecosystems. This preoccupation has lead the countries member of the Geneve Convention (trans-boundary pollution at long distance) to engage themselves for reducing the emissions of some of these metals and to encourage the survey of their deposition and the level of contamination in different compartments of the environment. Surveys of this type of pollution are under the influence of specific constraints because these compounds are present in trace concentrations. These constraints have led many countries to use bio-accumulating organisms, in which the pollutants are quantified. This biosurvey approach permits the evaluation of contamination levels and the exposure of living organisms to the effect of pollutants. Thus, starting with the beginning of 90s, a great number of countries are participating in a survey network of the atmospheric depositions of metals based on their dosage in terrestrial mosses, collected (where possible) in areas far from industrial pollution sources and far from the main circulation axes. Although the acute and chronic effects of heavy metals on humans may be well known, knowledge on the sanitary risks connected to their dispersion by aerial routes is still fragmentary. In the vicinity of some industrial emissions of heavy metals in the air (Pb, Hg and Cd) several studies have made evident a risk for the population in the close vicinity of these sources (Donisa et al., 2000). This risk is generally calculated on the basis of the available exposure data by modelling, considering the inhalation and the contamination of soils and water, which is transferred to humans through the food chain. For the long-range transport of pollutants, besides the possibility of calculating the risk, there are observations on their accumulation in the environment compartments and on the potential exposure of population in terms of remanence and their bioaccumulation.

364

R. Mocanu et al.

Heavy metals are non-degradable compounds and therefore they have the tendency to be accumulated in the biosphere. This is obvious, especially in north and central Europe where on a regional scale the contamination problems of some environmental compartments (forest humus, lakes and arable fields) arise from metal depositions as a consequence of long range atmospheric transport. The forest soils have a surface layer rich in organic matter. This type of soils seems to retain efficiently the atmospheric deposition of heavy metals. These soils are frequent in north and central Europe, where there are high concentrations of lead, cadmium, copper and zinc in the surface layer. Some authors estimate that these concentrations affect the biology of the soil, especially the decomposition of the organic matter, and that they have a baneful effect on the recycling of certain nutritive elements (Johansson and Sliggers, 1997). In several European regions the atmospheric contribution is significant for the contamination of agricultural soils (70 % of the Hg in The Netherlands, more than 90 % of Pb in the United Kingdom, 50 % of Cd in Sweden) (Johansson and Sliggers, 1997). The long range atmospheric contributions of several heavy metals during the last century, has led to the enrichment (by a factor of 5 to 50) of their concentration in the sediments of lakes in the north-eastern United States, southern Canada and southern Scandinavia. For Hg, for example, in southern Scandinavia and the northern United States this contribution has resulted in high concentrations in fish that are dangerous for human health and for fish and marine mammals (Johansson and Sliggers, 1997). Information on atmospheric heavy metals can be obtained by modelling and by measurements of actual atmospheric occurrences and/or deposition. Biomonitoring can be an efficient supplement and even replacement for these types of investigations, and also permits larger scale multiple sites programs, as has been shown by the running series of NORDIC moss survey (Rühling, 1994; Rühling and Steinnes, 1998). Biomonitoring may be defined as the use of bio-organisms/materials to obtain information on certain characteristics of the biosphere. With proper selection of organisms, the advantage of the biomonitoring approach is related primarily to the permanent and common occurrence of organisms in the field, even in remote areas, the ease of sampling, and the absence of any necessary expensive technical equipment. Mosses may be considered as the most commonly used biomonitor organisms. - Mosses do not have a well-developed root system and conducting system; the major source of the supply for essential nutrients (but not only) remains atmospheric deposition. - Mosses have a high capacity to retain many elements by surface sorption or by intracellular uptake. - Mosses retain and concentrate trace heavy metals. - Moss samples are easy to collect and also to analyze compared to precipitation samples. Mosses can be used for surveys of atmospheric depositions within a large as well as a smaller area. - They occur in almost all terrestrial ecosystems. - They have the ability to tolerate long periods of desiccation. This makes them able to colonize areas from north to south, with extreme environmental conditions. The main sources and mechanisms other than air pollution that influence the heavy metal content in mosses are: - Contributions from long-range atmospheric transport (Pb, As, Cd).

Heavy Metals Pollution: An Everlasting Problem -

365

Contributions from local sources (Cu, Cd, Co, Ni). Elements associated with local soil dust (Al, Cr, Fe, Sc). Transfer from soil via higher plants (Zn, Mo, Rb). Contributions from the marine environment (Mg, Mn, Na).

When considering the use of mosses as biomonitors, the active and the passive technique could be employed. The selection of one of these techniques depends mainly on the moss species, with respect to abundance, tolerance and their capacity for accumulating chemical elements. Extension of moss monitoring from the northern European countries (the starting areas for this technique) to more southerly areas, imposed the necessity of using epiphytic species instead of the epigeic ones. Another convenient alternative is to introduce active monitoring using moss transplants, in order to evaluate the exposure to contaminants and to calibrate the results obtained by monitoring with native moss species. This is the situation in lowland sites of Romania, where is difficult to find the classically known species for this purpose. In the eastern part of Romania (province of Moldavia) the most frequently found moss species is the epiphytic Hypnum cupressiforme. To investigate the compatibility of the two methods, passive and active monitoring, the epigeic moss Hylocomium splendens was used in transplants in parallel with epiphytically growing Hypnum cupressiforme employed for passive biomonitoring. The moss technique using the epiphytic Hypnum cupressiforme may be considered as a valuable method for estimating the atmospheric deposition of heavy metals in the catchment of River Prut (Cucu-Man et al., 2004). The data for atmospheric heavy metal deposition biomonitoring of the Prut River basin show rather uniform levels over the entire region, slightly higher in the same middle area of the river catchment (Figure 5), a similar trend as for water and sediments. Cd

Concentration

0.8

Element concentration in moss 50

0.4 0.2 0.0

Cu

Concentration (mg/kg)

0.6

0

Zn

40

1

2

3

4

5

6

7

8

9

S ite

30 20 10 0 0

1

2

3

4

5

6

7

8

9

Site

Figure 5.

Metal concentration in mosses collected in 2001 in Prut River catchment (mg·kg-1).

Considering the contribution of atmospheric deposition to the metal pollution of Prut River two ways of supply are possible: I. Direct deposition on the water surface. II. Deposition on land and supply to the river by surface runoff and snow melting.

R. Mocanu et al.

366

While the estimate of the latter needs information on factors such as surface soil enrichment of metal erosion rates, the former can be estimated roughly from the monitoring data. In order to use the information from the moss monitoring to estimate the atmospheric deposition in the area of concern the following assumptions have been made: a) Hypnum cupressiforme is structurally quite similar to Hylocomium splendens, therefore they may collect heavy metals to the same extent. The results obtained for the two species collected from the same sites show a significant correlation between the concentrations of typical pollutants (for example Pb, Figure 6). b) The calibration of metal concentration in Hylocomium splendens against bulk deposition from Norway (Berg et al., 1995) are valid also for the region concerned in the present study.

y = 2.1576x - 8.8584

Pb

2

R = 0.8019

50 40

Hc

30 20 10 0 0

Figure 6.

5

10

Hs

15

20

25

-1

Lead concentrations (mg·kg ) in Hypnum cupressiforme versus Hylocomium splendens.

Over the stretch of Prut River the Stânca Coste ti reservoir represents a major part of the total water surface (Figure 2a). Therefore the estimation was limited to the direct atmospheric 2 contribution to the surface of this reservoir, which is about 60 km . The values of heavy metal -1 concentrations in Hypnum cupressiforme (mg·kg ) moss samples from Stefanesti (the site closed to the Stânca reservoir) were selected. The yearly deposition values (mg·m -2·y-1) were calculated. Assuming the mean depth of the reservoir to be 12.5 m, the total volume is of the 8 3 order of 7,5u10 m . The annual atmospheric addition of metals to the reservoir were calculated from the estimated deposition rates. If diluted over the whole volume, and given that the residence time of water in the reservoir is about 4 month, the increments to the metal concentrations in water are estimated Comparing the obtained values with those from the monitoring data of the Prut River water shows that the contribution from direct atmospheric deposition to the reservoir is less than 5 %, and hence small compared to contributions from other sources.

Heavy Metals Pollution: An Everlasting Problem

367

Concentration in Hypnum cupressiforme (mg·kg-1)

Cu – 11.8 Zn – 28.9 Cd – 0.42 Pb – 14.4

Yearly deposition values -2 -1 (mg·m ·y )

Cu – 3.0 Zn – 9.0 Cd – 0.17 Pb – 4.5

Annual atmospheric addition of metals to the reservoir (kg)

Cu – 200 Zn – 500 Cd – 10 Pb – 300

Atmospheric contribution to the concentrations in water (µg·L-1)

Cu – 0.09 Zn – 0.22 Cd – 0.005 Pb – 0.15

Contribution from direct atmospheric deposition to the reservoir (%)

Concentrations in water (µg·L-1)

Cu – 10 Zn – 20 Cd – 0.5 Pb – 3

Cu – 1 Zn – 1 Cd – 1 Pb – 5

The obtained values for atmospheric deposition in the Prut River catchment are lower than those obtained in Transilvania (Stan et al., 2001), similar to those obtained in the Eastern Carpathians and higher compared to the Republic Moldova (Figure 7). This supports the conclusion that no trans-boundary pollution of the investigated elements from Romania to the Republic of Moldova could be considered. Conclusion The above results lead to the conclusion as far as the heavy metals behavior and transport are concerned that Romania is still an importer and an interjacent of pollution but is not an important producer itself (Figure 7).

R. Mocanu et al.

368

Pb Concentration (mg/kg)

Concentration (mg/kg)

15

10

5

200

100

0

0 Transilvania

Eastern Carpathians

Prut catchment

Transilvania

Moldova

Cd

Concentration (mg/kg)

Eastern Carpathians

Prut catchment

Moldova

Cu

200

4.0 Concentration (mg/kg)

Zn

300

20

150

3.0

100

2.0 1.0 0.0

50

0

Transilvania

Figure 7.

Eastern Carpathians

Prut catchment

Moldova

Transilvania

Eastern Carpathians

Prut catchment

Moldova

Heavy metal concentrations (mg·kg-1) in mosses collected in Transilvania, Eastern Romania, Prut River catchment and Republic of Moldova.

References Berg, T., O. Røyset and E. Steinnes; Moss ( Hylocomium splendens) used as biomonitor of atmospheric trace element deposition: estimation of uptake efficiencies, Atmos. Environ. 29 (3) (1995) 353-360. Cucu-Man, S., R. Mocanu, O. Culicov, E. Steinnes and M. Frontasyeva; Atmospheric deposition of metals in Romania studied by biomonitoring using the epiphytic moss Hypnum Cupressiforme, Int. J. Env. Anal. Chem. 84 (11) (2004) 845-854. Donisa, C., R. Mocanu and E. Steinnes; Heavy metals pollution by atmospheric transport in natural soils from North Eastern Carpathians, Water Air Soil Pollut. 120 (2000) 347-358. Johansson, K. and J. Sliggers; Overview –Effects of heavy metals, Workshop on critical limits and effects based approaches for heavy metals and persistent organic pollutants, Bad Harzburg, Germany (1997). Rühling, Å. (Ed.); Atmospheric heavy metal deposition in Europe. Estimations based on moss analysis, NORD 1994:9, Nordic Council of Ministers, Copenhagen (1994). Rühling, Å. and E. Steinnes (Eds.); Atmospheric heavy metal deposition in Northern Europe. 1995-1996, NORD 1998:15, Nordic Council of Ministers, Copenhagen (1998). Stan, O., A. Lucaciu, M.V. Frontasyeva and E. Steinnes, New results from air pollution studies in Romania, Proc. NATO Advanced Research Workshop on Monitoring and Man-Made Radionuclides and Heavy Metals Waste in Environment, Kluwer Academic (2001) 179-190.

Atmospheric Wet Deposition Monitoring in Iasi, Romania Cecilia Arsene1, Nikos Mihalopoulos2, Romeo-Iulian Olariu1, and Marius Duncianu1 1

“Al.I. Cuza” University of Iasi, Faculty of Chemistry, Analytical Chemistry Department, 700506 Iasi, Romania 2 Environmental Chemical Process Laboratory, Department of Chemistry, University of Crete, 71409 Heraklion, Greece Key Words: Wet Deposition, Monitoring, Environment, Precipitation, Fluxes, Iasi

Introduction The degradation of environmental quality in many regions of the world has accelerated during the past decades especially due to the industrial development, which has led to important changes in different compartments of the environment. Important atmospheric species are considered to be responsible for wide spread environmental effects including, changes in pH deposition, corrosion of buildings material etc. Deposition of air pollutants is an important loss process for most of the species present in the atmosphere that can cause severe damage to ecosystems. Air pollutants are deposited to the earth’s surface especially through wet and dry processes. Deposition rates are determined in order to estimate the impact of these pollutants on ecological systems. Deposition of pollutants by wet processes is relatively easy to determine through analysis of precipitation samples. However, it is well recognised that less is known about dry deposition, which is much more difficult to measure (estimated using measured air concentration and the deposition velocity concept) and which appears to predominate near strong emission sources with wet deposition predominating further downwind (Whelpdale et al., 1997). Direct measurement of pollutants deposition by dry processes is more difficult and requires extensive instrumentation and technical resources. A combination of modelling and measurement activities has been developed in order to achieve effective emission reductions or in order to monitor the progress in terms of deposition (Erisman et al., 2001). While in America and Western Europe the measurement of air pollution has become practically a routine matter for Eastern European countries this is a present-day task. There are many observations in the literature that refer to the importance of wet deposition analysis, in most of Europe, North America and Asia, for obtaining information on the contribution of industrial activities to atmospheric deposition (Agrawal and Singh, 2001; Gao et al., 2001; Nilles and Conley, 2001). Within the last decades, due to intense research studies, acid deposition is now a reasonably well-understood environmental phenomenon. Many researchers blame acid precipitation for much of the damage done to trees and buildings as well as the rising number of people suffering from respiratory problems. Many data concerning acid deposition phenomenon in several regions of the world have been published but a summary for the whole globe does not exist. It has been shown that regions with the highest precipitation concentrations and deposition fluxes of sulphate and nitrate coincide closely with regions having the highest density of SO2 and NOx (NO + NO2) precursor emissions and that this is occurring primarily in the regions where a large fraction of the world’s fossil fuel is consumed (Whelpdale et al., 1997). Presently it is well established that the chemical composition of wet precipitation is strongly affected by the chemical composition of the atmosphere, and is especially affected by the source zone traversed by the cloud system. Particles and gases either incorporated directly 369 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 369–377. © 2006 Springer. Printed in the Netherlands.

370

C. Arsene et al.

in the cloud (rainout), or/and washed out by the rain droplets below the cloud (washout) are directly responsible for the chemical species identified in rain. It is considered that exchanges between the two phases, rain and aerosol, account for the diversity observed in rain composition. These effects depend on the meteorological, as well as the microphysical conditions (Choularton and Bower, 2001; Samara and Tsitouridou, 2000 and references therein). Although in most regions of the world increased attention has been given to atmospheric deposition in order to elucidate the effect of different types of emission both on dry and wet deposition, for Romania this is a new and contemporary task. The present study, which is a first step in a larger monitoring program meant to be developed in the future in Romania, is focussed on determining the elemental composition of wet deposition and in evaluating the influence of different sources of pollution types on the rainwater pH. In order to assess the main task of the study, wet depositions were collected over a year from May 2003 to May 2004, in Iasi, the biggest city in the Moldavian region and one of the largest cities in Romania with approximately 350,000 inhabitants. The city lies in the Northeast part of Romania and is surrounded by seven hills. As for the climate, due to the geographical location, Iasi city is characterised by a o excessive continental climate with high differences of temperature from winter (-20 –0 C) to o the summer season (20 – 35 C). Unfortunately the sampling site is located in a region that may be potentially largely affected by urban or industrial emissions. Therefore, the deposition values can be perceived as upper limits of what might be considered as regionally representative values. This will be an advantage at a later stage when an enlargement of the database is expected, i.e. a database containing also results from less polluted regions (rural or remote areas). Over a year, however, the precipitation height was about 40 cm. The measured pH values of the rainwater varied in the range of 5.0 to 7.0 for most of the samples. During the winter only January 2004 and partially February 2004 were mostly characterised by snow events, and the pH values of the melted snow water varied in the range of 4.5 to 5.0. The obtained data, however, allow us to make a classification of the rain events in different categories according to their origin, to develop a climatology of air masses associated with rain events, to establish a dependence of pH on the air mass origins, to study the insoluble + fraction as a function of pH and air mass origin, and also to estimate the fluxes of [H ] and the insoluble fraction. Chemical characterisation of both the soluble and insoluble fractions is underway. Data collection Geographical and meteorological characteristics of the sampling site The monitored site lies between 47°20'' northern latitude and 27°60'' eastern longitude (Figure 1). At this latitude incident solar zenith angles between the solar radiation and the Earth’s surface of about 19°23' during the winter and 66°17' during the summer, at midday, are reached with a variation of about 47°. The warmest month of the year is July with a mean monthly temperature of about 24 °C whereas January is the coldest with a mean monthly temperature of about 3 °C. An annual o mean temperature variation of about 28 C is estimated, which classifies Iasi as a region with very a high thermic amplitude corresponding to a severe continental climate. During winter and summer a small difference of the air temperature is observed from one month to another

Atmospheric Wet Deposition Monitoring in Iasi, Romania

371

within the same season and during spring and autumn this difference can be as high as 10 °C. Usually, at the end of autumn and the beginning of spring within a time range of 12 hours a large difference in the air temperature may occur with important negative consequences for population, agriculture and buildings. Due to its location Iasi city is directly exposed to winds coming from a wide range of directions and hence to the corresponding atmospheric fronts. Winds from the west (W), southwest (SW), southeast (SE) often prevail in the area. There are also situations when the wind direction is from the east (E), northeast (NE) and northwest (NW). During north-easterly winds the sampling site can be regarded as representative of the transfrontalier atmosphere (air fronts coming from the former Soviet Union) while during westerly and south-westerly winds it can be regarded as representative of the urban atmosphere (including industrial activities, agriculture land activities).

Figure 1.

Sampling location and Iasi region.

The topography of the city is dominated by the presence of an important number of hills due to which a local dynamics of the atmospheric air fronts is observed. The difference o in the temperature values measured at the bottom and tops of the hills (at least 1 or 2 C colder at the top of the hill) is also responsible for the appearance of this phenomenon. Method for sampling wet deposition The sampling point was located on the roof of the University building (approximately 20 m from the ground) which is beside the emerging top of the Copou hill of Iasi city. The sampling system was raised 2 m above the roof before precipitation events started. Generally, precautions were taken in order to avoid exposure of the funnel to dry contamination. Rainwater samples were collected on an event basis by using a simple self-developed sampling system (Figure 2). Where possible, rainwater sampling system parts made of polyethylene, polypropylene and 5 litre capacity containers made also of polyethylene terephtalate were used for sampling. The materials employed during the sampling procedure (funnels, 5 liter capacity containers, additional vessels used to measure the volume) were well washed each

372

C. Arsene et al.

time with liquid cleaning soap, rinsed with a solution of nitric acid and than washed often with deionised water. Dry contamination on the collector funnels was minimized by rinsing funnels with distilled water, as close as possible to the start of rainfall. When rainfall began in the middle of the night the exposure of the funnel to dry contamination may have been for a longer time. The regular cleaning procedures and the nature of the site meant that contamination from insects or birds was not a significant problem. Precipitation samples were collected from the field site as soon as possible after the cessation of rainfall and filtered through Whatman no 41 filter paper. A fraction of the rain samples was preserved immediately with chloroform (CHCl3) o and refrigerated at 4 C for later chemical composition analysis. The pH was measured with a pH meter OP-401/2, Radelkis, using a glass electrode, standardised with pH = 6 buffer solution. The hydrogen ion concentration was calculated from the measured pH. Presently the samples have not yet been completely analysed from the chemical composition point of view. Although the equivalent conductivities of the major ions were not measured it is known that these are roughly equal (with the exception of hydrogen). In fact conductivity can be used to estimate the total amount of dissolved ionizable material incorporated by scavenging processes. For further samples conductivity will be measured and as it varies with temperature the obtained results will be corrected for a standard temperature. Figure 2.

The own developed sampling system.

The pH of a solution is defined in terms of the negative logarithm (to base 10) of the hydrogen ion activity (aH) by the following expression: pH = -log10 (aH) For rainwater, the hydrogen ion activity is a good approximation of the hydrogen ion + + concentration, [H ], therefore it is possible to estimate [H ] from pH measurements by the expression: +

[H ] = 10

-pH

×10 6 (Peq L-1)

A parameter used to characterize rainwater is OH+ that represents the equivalent conductivity of H+ estimated to be 0.350 (PS cm-1)/(Peq L-1). The major soluble component of the atmosphere is CO2 but this is only weakly ionized in solution and the carbonic acid (H2CO3) generated defines a lower limit for background pH of 5.6 at 15 oC (Beverland et al., 1997 and references therein). As Beverland et al (1997) show, although a pH increase of 0.061, due to

Atmospheric Wet Deposition Monitoring in Iasi, Romania

373

decreasing Henry’s Law constant, can be expected between 5 and 25 oC for carbonic acid alone, changes of CO2 equilibrium cannot be responsible for the observed discrepancies of rain samples collected within the pH range 3.9 to 4.9. The major species which contribute to an increased acidification of rainwater are SO2 and sulphuric (H2SO4), nitric (HNO3) and hydrochloric (HCl) acids. Results and discussion Information (not fully complete) regarding a real assignment of wind direction is available for the previous year. Dependence between rain and air mass climatologies has been followed and it illustrates a highly uncertain trend. However, a northwesterly prevalent direction of air mass trajectories for almost all wind directions has been observed (Figure 3). A view of the pH and temperature trends over the year is presented in Figure 4. However, in the analysis of the data seasonal trends were closely followed and this allowed us to decide if meaningful changes in the observed parameters from one season to another occurred. The observed difference in the values of pH, with smaller values especially during the winter season, probably originates from combustion sources that include a considerable seasonal contribution from heating. The study further of pH indicates that for the summer and + autumn of 2003 rainfalls were more alkaline with less H ions, but as these seasons preceded + precipitation showed lower pH and higher H ion concentrations (Figure 5). Probably in this period the particulate load in the atmosphere declined substantially and hence aerosol rich in + basic ions declined leading to an increase in the H ion concentration and a subsequent lowering of rainfall pH.

Figure 3.

Prevalent direction of air masses trajectories for almost all wind directions.

As it was previously shown most of the rainwater has a pH of 5.6 to 5.8 due to the presence of carbonic acid (H2CO3) formed from dissolved atmospheric CO2 gas and H2O. Precipitations with values of pH below 5.6 or so are considered acid rains, while those with values above are basic rains. The data on mean monthly pH values of wet deposition indicate that rainwater in the Iasi area is to some extend acid, basic or normal accordingly to some remarks made by a specialist with respect as to what represents neutral rainwater. When water vapour reaches the atmosphere values for a non-acidic system is available (pH 7) but, however, in the atmosphere due to a lot of both natural and non-natural sources of materials deviation from the neutral value (5.6 - 5.8) of the rain pH can occur.

C. Arsene et al.

374 305

8.00

40

8.00

300 7.00

7.00

295

275

4.00

) -1

6.00 20 5.00

[H+] ( P echiv L

280

Units of pH

Units of pH

285 5.00

Temperature (K)

30 290

6.00

10

270

4.00

3.00 265 2.00 29-Apr- 18-Jun- 07-Aug- 26-Sep- 15-Nov- 04-Jan- 23-Feb- 13-Apr03 03 03 03 03 04 04 04

260

Day of the year (DD:MM:YYYY)

3.00 29-Apr- 18-Jun- 07-Aug- 26-Sep- 15-Nov- 04-Jan- 23-Feb- 13-Apr03 03 03 03 03 04 04 04 Day of the year (DD:MM:YYYY)

0

Figure 4. Variation of pH and temperature over the Figure 5. Variation of pH and H+ ion concenyear at the monitored site (open symbols tration over the year at the monitored + – pH and full symbols – temperature). site (circle –pH and square –[H ]).

Table 1. Number of the precipitation events over the year, monthly mean values of the + pH, H ion concentrations and the total monthly amount of rainfall. No. of precipitation Year Months events rain snow I E May II N-W 2 I W 2 June II N 2 I N-W 7 July II E 4 I N-W 2 August II W 2 2003 I E 3 September II S-W I N 2 October II N 4 I E 2 November II E 1 I W 2 December II S-W 1 1 I W 1 3 January II S-W 2 I N 2 1 February II S-W 1 2004 I W 1 1 March II N-W 2 I N-E 1 April II W I –first decade (beginning till middle of the month) II –second decade (middle till end of the month). Trajectory wind direction

Total monthly amount of rain

+

pH

[H ] -1

5.47 5.72 6.20 6.15 6.27 5.42 6.20 5.85 5.70 5.73 4.87 5.55 5.30 5.55 4.77 5.02 5.08 5.43 6.22 6.90 6.65 -

(Peq L ) 3.35 2.14 0.64 0.91 0.66 4.06 0.64 1.30 2.11 2.29 20.12 2.81 6.72 4.22 19.21 9.45 10.41 3.54 0.67 0.13 0.22 -

(mm) 28.88 15.77

32.51

36.39 38.62

84.89

5.67 25.13

32.35

24.26 38.56 23.83

The total number of the precipitations over the year was 52, with 43 rain events and 9 snow events. A total rain column of 405 mm from May 2003 to May 2004 was measured. The + monthly mean values of the pH and H concentrations of rain given by a mean of the number of rain events are presented in Table 1.

Atmospheric Wet Deposition Monitoring in Iasi, Romania

375

Air masses trajectories were obtained using HYSPLIT: On-line Transport and Dispersion Model (Draxler and Rolph, 2003; Rolph, 2003). An example of the calculated air front trajectories is shown in Figure 6. Rain events were classified in different categories according to their origin, as derived from 120 h air back mass trajectories based on 1000 and 3000 m as for the altitude. From the analysis of these trajectories westerly air fronts appear to be the most frequent (Figure 7). The inset in Figure 7 shows the percentage of precipitation events associated with the air masses fronts. Northwesterly and westerly air fronts appear to induce the most abundant precipitation events over the year, especially during summer and winter. Figure 6. Example of the calculated backward air mass trajectories.

Figure 8 shows the distribution of pH as a function of the air mass origin. From the observed behaviour in Figure 8 it is clear that there are only a few precipitation events with pH values between 4 and 5 while for the others the pH values are in the range of 5 –6 or 6 –7 (about half).

Figure 7. Percentage of air masses trajectories over the year and of the precipitation events.

Figure 8. Distribution of pH as a function of the air mass origin.

From the distribution of the insoluble fraction masses versus pH it is observed that a dependence exists but with a low correlation coefficient (Figure 9). However, a better correlation coefficient is obtained when the mass concentration of the insoluble fraction is plotted as a function of pH (Figure 10). A statistical treatment of the data over a 0.5 unit of pH range leads even to a better correlation coefficient for the dependence of the insoluble fraction mass concentration and pH (0.88 compared to 0.57).

376

C. Arsene et al.

Figure 9. Distribution of insoluble part masses versus pH.

Figure 10. Insoluble part mass concentration as a function of pH.

From an analysis of the data in Figures 8 and 9 it appears that the pH values increase with increases of the insoluble fraction. It is well known that the pH of the aqueous environment is one key property that defines the nature of the chemical species. Thus, the increase in the insoluble mass fraction with increase of the pH rain droplets could indicate that the insoluble matter may contain in its chemical composition significant amounts of terrigenic elements (e.g. iron oxides) that become more insoluble in rain droplets as the pH values increase. It might be as well that compounds are present whose solubility increases with decreasing pH. The concentration of the insoluble fraction is variable for most of the air masses with more intense variations for north-westerly and northerly trajectories (Figure 11).

Figure 11.

Concentration ranges of the insoluble part at the monitored site as a function of the air mass origin.

Fluxes either of the H+ ion concentration or of the insoluble fraction were calculated from their concentration and the volume of the rainfall during the sampling at the monitored + site. Both H and insoluble fraction fluxes follow almost the same trend over the year with small values up to the end of 2003 and with higher values at the beginning of 2004. Observations In 2003, a wet-deposition collection site was established in Iasi, northeast Romania, to monitor the quantity and chemical quality of atmospheric wet deposition. Monitoring was carried out from May 2003 to May 2004. During the first phase of sampling 52 wetdeposition samples were collected. A second phase of sampling began in May 2004 and is

Atmospheric Wet Deposition Monitoring in Iasi, Romania

377

still on-going. Wet-deposition samples collected during the first phase of sampling were first analyzed for pH. Meteorological aspects were also closely followed. Presently no chemical speciation concerning the chemical composition of the wet deposition is available. Precipitation caused by air masses originating in west and northwest Europe (48% of total rain events) were characterised by the highest concentration of the insoluble fraction. However, after a complete analysis of the samples from the chemical composition point of view, closely followed by a careful analysis of the meteorological parameters, attempts will be made to obtain full information regarding the behaviour of all the measured pollutants in the atmosphere. Although the presented activity has already begun to produce useful information, its real benefit will be realized after it has been in operational mode for several years. This will enable the examination of the trends of deposition and atmospheric acidity, in particular in relation to the emissions of regulated pollutants by national and international laws. Acknowledgment Financial support from NATO EST.ARW 980164 action for travel funding is gratefully acknowledged. The authors gratefully acknowledge the NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and dispersion model and/or READY website (http://www.arl.noaa.gov/ready.html) used in this publication. References Agrawal, M. and R.K. Singh; Effect of industrial emission on atmospheric wet deposition, Water, Air, and Soil Pollut., 130 (2001) 481-486. Beverland, I.J., M.R. Heal, J.M. Crowther and M.S.N. Srinivas; Real time measurement and interpretation of the conductivity and pH of precipitation samples, Water, Air, and Soil Pollut., 98 (1997) 325-344. Choularton, T.W. and K.N. Bower; The role of clouds processing in the relationship between wet deposited sulphur and sulphur dioxide emissions, Water, Air, and Soil Pollut.: Focus, 1 (2001), 365-372. Draxler, R.R. and Rolph, G.D., 2003. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model access via NOAA ARL READY Website (http://www.arl.noaa.gov/ready/hysplit4.html). NOAA Air Resources Laboratory, Silver Spring, MD. Erisman, J.W., A. Hensen, D. Fowler, C.R. Flechard, A. Gruner, G. Spindler, J.H. Duyzer, H. Weststrate, F. Romer, A.W. Vonk and H.V. Jaarsveld; Dry deposition monitoring in Europe, Water, Air and Soil Pollut.: Focus, 1 (2001) 17-27. Gao, S., K. Sakamoto, D. Zhao, D. Zhang, X. Dong, S. Hatakeyama; Studies on atmospheric pollution, acid rain and emission control for their precursors in Chongqing, China, Water, Air, and Soil Pollut., 130 (2001) 247-252. Nilles, M.A. and B.E. Conley; Changes in the chemistry of precipitation in the United States, 1981-1998, Water, Air, and Soil Pollut., 130 (2001) 409-414. Rolph, G.D., 2003. Real-time Environmental Applications and Display sYstem (READY) Website (http://www.arl.noaa.gov/ready/hysplit4.html). NOAA Air Resources Laboratory, Silver Spring, MD. Samara,C. and R. Tsitouridou; Fine and coarse ionic aerosol components in relation to wet and dry deposition, Water, Air, and Soil Pollut., 120 (2000) 71-78. Whelpdale, D.M., P.W. Summers and E. Sanhueza; A global overview of atmospheric acid deposition fluxes, Environ. Monit. Assess., 48 (1997) 217-247.

Problems of Air Quality in Tashkent City G. A. Tolkacheva Research Hydrometeorological Institute, Uzbekistan, Tashkent, K. Makhsumov st., 72 Key Words: Air pollution, Carbon dioxide, Metals, Ozone, Precipitation (wet and dry), Sulphur dioxide, Toxicity

Introduction During last years the negative effects of anthropogenic pollution on the environment has intensified. The effects are observed on the regional and local levels in the change of the environment quality, as well as on the global level as in the change of the planet climate. Urban agglomerations play an important role in these processes. The interconnection between the problems of environment, economical development and demographical processes is sensed in urban agglomerations to the highest extent. In a big city, a specific natural technogenic ecosystem is formed, with new trends in biological processes of exchanging the energy and chemical substances. These processes determine, to a certain extent, the urban environment quality. Atmospheric air is the most agile component of the environment. It is subject to the impact of emissions; under their effect various chemical processes occur in the atmosphere, which result in formation of secondary, often more toxic trace components. In this regard, the assessment of the air quality in big industrial urban areas is a very timely and actual task. Tashkent city is the capital of the Republic of Uzbekistan, and is the largest city in Central Asia. The city area covers 259.39 km2 with a population of 2121 thousand. The city is situated in the western part of Central Asia, on a wide foothill plain on the right-hand bank of the Chirchik river, at 440-480 m a.s.l. The eastern part of the city is slightly hilly while the western part is located on a plain. The climate of this territory is extremely continental and dry, with hot summers and relatively cold winters. The area has a high climatic potential for atmosphere pollution. The following aims and tasks for studies in Tashkent city as an exemplary case have been untertaken: x

Assessment of the present state of the atmospheric pollution in Tashkent with the data from systematic observations;

x

Derivation of the relationship between the levels of atmospheric pollution and industrial emissions;

x

Evaluation of the role of chemical composition of the atmospheric fallouts (dry precipitation) as the indicators of atmospheric pollution;

x

Simulation of the effect of atmospheric phytotoxic components on the land vegetation; 379

I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 379–392. © 2006 Springer. Printed in the Netherlands.

G. A. Tolkacheva

380 x

Derivation of the criteria for the evaluation of the effect of the polluted air on vegetation, and of the effect of atmospheric precipitation on the environment;

x

Definition of the main directions for the improvement of the air quality system.

Objects and Methodology of Studies Objects such as atmospheric air, atmospheric fallouts (dry and wet precipitation) and land vegetation in Tashkent were chosen for the studies. Systematic observations of the quality of atmospheric air in Tashkent are carried out at 13 stationary observation points conforming to the RD 52.04.186-89 standard (Figure 1).

Figure 1.

Map of Tashkent city.

Samples are taken at least 4 times a day. The data from measurements of the main trace components are compared by the MAɋone time and MACav.daily for the detection of causes and peculiar features. In urban air pollution, the information about climatic conditions determining the transfer and dissipation of trace components in the atmosphere is used along with the

Problems of Air Quality in Tashkent City

381

information about the amount of harmful substances emissions from industrial plants and traffic. Measurements of the concentrations of suspended substances, sulphur dioxide, nitrogen oxides and carbon oxide in the urban atmosphere are obligatory. The index of atmospheric pollution (IAP) is used for estimation of the integrated atmospheric pollution by an array of pollutants: Ji

¦( i 6

qi )c i MACav.daily

where: qi is the mean annual concentration of the i-th substance, MACi is the mean daily maximum admissible concentration (MAC) of the i-th substance and ci is the factor describing the degree of harmfulness of the pollutant (0.85-1.5). Observations of the chemical composition of atmospheric fallouts (dry precipitation) have been carried out in Tashkent since the 80s of the last century, in accordance with standardization procedures worked out at SANIGMI. Chemical analyses of the atmospheric air samples and atmospheric fallouts were made using the methods of atomic adsorption, ion chromatography, photo-colorimetry, etc. The methodology of studying the effects of trace components in simulation chambers was developed for the assessment of the impact of phytotoxins on land vegetation. On the basis of the experimental studies in chambers and of the field studies, ecological standards have been formulated for the effect of atmospheric air pollution on land vegetation. Discussion of Results and Evaluation of the Current State of the Atmospheric Air Pollution in Tashkent city Table 1 presents the dynamics of the atmospheric air pollution as established from the data on systematic observations of the main pollutants in MAC units. It was established that in general, the mean long-term levels of air pollutions exceed the admissible values for the following trace components: dust, nitrogen oxides and ozone. This are a number of factors responsible for this. Ɍɚble 1. Pollutant Ozone Dust Nitrogen dioxide Carbon oxide (II) Nitrogen oxide Sulphur dioxide

Dynamics of the atmospheric air pollution according to the main pollutants in measured in MAC units. 91 92 93 0.8 2 2 1.33 1.33 1.33

94 2 2.0

Years (1991-2002) 95 96 97 98 1.3 2.3 1.3 1.3 2.67 2.0 2.0 2.0

99 00 1.3 2.53 1.33 1.33

01 0.67 2.0

02 1.06 2.0

1.5

1.0

1.0

1.0

1.0

1.0

1.0

1.5

1.75

1.75

1.5

2.0

1.0

1.33

1.0

1.0

1.0

1.0

1.0

0.66

0.66

0.33

0.33

0.66

0.83

0.5

0.5

0.5

0.3

0.5

0.5

1.0

1.0

0.83

0.66

0.83

0.2

0.16

0.2

0.2

0.2

0.12

0.34

0.26

0.2

0.2

0.2

0.24

382

G. A. Tolkacheva

The pollution with dust is caused by emissions from industrial plants and traffic, as well as by high emissions from natural sources of dust. The main emission sources of nitrogen oxides are mainly the mobile ones such as motor vehicles. The relatively high level of pollution by ozone is determined by photo-chemical reactions in the atmosphere under the influence of solar radiation and the high level of ozone precursor compounds. Figure 2 presents the zones most strongly affected by the emissions from industrial plants and the local zones with increased atmospheric pollution. It is evident that the field of concentration of pollutants over the city is formed by the emissions. The analysis of the dynamics of the emissions from stationary sources in Tashkent during the last years is shown in Figure 3. The graph of the dynamics of emissions from stationary sources during the last 10 years has also been plotted to reveal possible relationships.

Figure 2.

Industrial zones in Tashkent and local zones of increased atmospheric pollution.

Problems of Air Quality in Tashkent City

Figure 3.

383

The dynamics of emissions from stationary sources in Tashkent for 1992-2001.

Figure 3 shows that during the last ten years the emission of pollutants from stationary sources in the city has been significantly reduced. Carbon oxide accounts for more than half of the overall pollutant emissions. Components such as nitrogen oxides, solid substances and hydrocarbons make a significant contribution to the overall emissions. During the last 10 years the share of hydrocarbons in the emissions has increased significantly. The increase in the average concentrations of dust and nitrogen oxides in the urban air can result from the motor-transport emissions, which account for about 90% of the total emission of the pollutants. Analysis of the seasonal dynamics of the change in the long-term values of the pollutant concentrations sampled at different sampling points has revealed the local peculiar features of the air pollution by the different trace components. Figure 4 presents mean long-term seasonal trend of the change in dust concentrations. The highest concentration values were recorded at the measurement point located in the area subjected to the solid emissions from the Tashkent heat-producing power station. In this case an increase in the dust concentration during summer and autumn months is observed.

Figure 4.

Annual trend of dust concentration (mg/m3) at individual measurement points (average values for 1991-2000).

384

G. A. Tolkacheva

The concentrations of nitrogen dioxide in the air measured at all sampling points were almost the same, so the annual trend is negligible (Figure 5).

Figure 5.

Annual trend of nitrogen dioxide concentrations (mg/m3) at individual measurement points (average for 1991-2000).

The situation is more complicated with respect to the levels of carbon oxide (II) and its annual dynamics observed at individual points (Figure 6). A spread of concentrations is observed depending on the season and on the location of each measuring point.

Figure 6.

Annual trend of carbon oxide concentrations (mg/m3) at individual measurement points (average for 1991-2000).

The dynamics of the seasonal variation of the ozone concentrations is presented in Figure 7, for 10 years of observations. It has been established that, during almost all the years of observations, the maximum concentrations were measured in summer months: June, July, and August. When comparing the observed values of the ozone concentration in the surface atmosphere layer recorded in Tashkent and Central Europe, it was revealed that the concentration ranges are very similar and fall in the range of 0.015 – 0.085 mg/m3. A complex analysis of the systematic data on atmospheric air pollution in Tashkent and of the development in the industrial structure and traffic has made it possible to assess in general the spread of the technological stress over the city territory (Figures 1, 2).

Problems of Air Quality in Tashkent City

Figure 7.

385

Dynamics of change of the surface ozone concentrations in Tashkent by months (1990-1999).

Assessment of the Role of Chemical Composition of Atmospheric Fallouts (dry precipitation) as an Indicator of Atmospheric Pollution In big industrial urban agglomerations, such as Tashkent in particular, an important factor concerning the effect of atmospheric pollution on the environment is the mechanism of removal of trace components from the active biological cycles. They are subdivided into classes where the following mechanisms and processes are involved: x

Chemical mechanism of transformation determined by the reactions of oxidation, hydrolysis and photochemical conversion;

x

Biological decomposition which is largely determined by intermediate microbiological reactions which take place in soil, water sediments and atmospheric aerosols;

x

Physical mechanisms, dissolution, agglomeration and gravitational deposition (sinking).

The main mechanisms of the removal of suspended particles and gases from the atmosphere are physical ones: dry and wet deposition or fallout in the form of dry and wet precipitation. Wet atmospheric precipitation comprises rain, snow and hail. Dry atmospheric fallout is the gravitational deposition (sinking) of dust particles and gas ingredients which are deposited on and adsorbed by the objects comprising the environmental surface. Dry fallout is the main mechanism of the removal of particles from troposphere below 100 m. It includes two processes: gravitational deposition and particle accumulation on surfaces of objects (buildings, plants, soil, etc.). The effect of the removal of particles from

G. A. Tolkacheva

386

the atmosphere is closely related to physical changes of aerosol particles (increase of particle size with time, coagulation and other processes). The effect of deposition of gas components is determined by their reactivity, by the concentration of polluting components, solar radiation intensity and the oxidizing potential of the atmosphere. Under the conditions of the dry arid climate of Tashkent, the dry deposition of pollutants prevails over wet deposition during the entire year. Information about dry atmospheric fallouts (DAF) is used for the estimation of integral characteristics of urban atmosphere pollution, in order to reveal the ways of transformation of trace components released from the emission sources. Based on the results of the long-term systematic observations of the density of DAFflows, a chart-plot was constructed for the Republic of Uzbekistan. The map thus produced shows that Tashkent is located within the zone where the overall density of DAF-flow is 500 – 1000 kg/ha. Certain chemical transformations take place in the atmosphere as the result of interaction between oxygen, moisture, solar radiation and sulphur dioxide, nitrogen oxides and carbon oxide as follows: hQ SO 2  2 OH o H 2SO 4 SO 2  H 2 O o SO 2 ˜ H 2 O SO 2 ˜ H 2 O o HSO 3-  H  NO 2  OH  M o HNO 3  M N 2 O 5  H 2 O o 2 HNO 3 NO 2  NO 3  M o N 2 O 5  M N 2 O 5  H 2 O o 2 HNO 3 PhOH  NO 3 o PhOH  HNO 3 CO  O o CO 2 ­CO CO 2  HOH œ H 2 CO 3 o ® 2 ¯H 2O

End products of these transformations are sulphates, nitrates, carbonates and hydrocarbonates. These components can deposit on the surface of aerosol particles, and contribute to DAF composition and impact. Figure 8 presents pie diagrams of the averaged data on the mineral composition of the liquid extracts from the dry atmospheric fallouts sampled at Tashkent meteostation in the centre of the city, at Almalyk meteostation in the outskirts of the town and, for comparison, similar data from the meteostations “Chatkal Nature Reservation” and “Abramov glacier”. It is necessary to note that the composition of dry fallouts sampled in Chatkal with the northwestern wind direction (which prevails in that area) can be influenced by the transfer of sulphur dioxide and other components from Almalyk town which is the major source of anthropogenic pollution in Tashkent province. The possibility of such atmospheric transfer has also been confirmed by model calculations.

Problems of Air Quality in Tashkent City

Figure 8.

387

Relationships between the water-soluble mineral components in dry atmospheric fallouts at various sampling points.

Figure 8 shows that each sampling point is characterized by a certain relationship between water-soluble components and their content in a liquid extract. The lowest mineralization was detected in samples taken from dry fallouts at the meteostation “Abramov glacier”, while the highest was detected in dry fallouts sampled at the Almalyk meteostation. In all samples, except those from the Abramov glacier, the prevalence of hydrocarbonates was observed, as they are, firstly, the main soil component, and secondly, the product of carbon oxide (which is the main emission from traffic) transformation into hydrocarbonates. The highest content of sulphides is typical for the anthropogenic aerosol. The maximum sulphide concentrations were recorded in Almalyk.

Figure 9.

Concentration of solid particles in the atmospheres of towns with population of more than 100 000 people (data from the Hydrometeorological service).

G. A. Tolkacheva

388

Figure 10.

Total content of PAH in atmospheric air of industrial towns in Uzbekistan.

Figure 9 presents data on the dynamics of the dust concentration in the atmosphere of several urban areas, including Tashkent. Figure 10 presents the data on the total concentration of the polyaromatic hydrocarbons (PAH) in the atmosphere of several industrial towns, including Tashkent. It is known that stable organic substances, including PAH, are concentrated in the atmosphere on fine-disperse dust particles. The results of these studies have confirmed this fact. Atmospheric Precipitation – Indicator of the Assessment of Climatic and Chemical Relationships The chemical content of the monthly total precipitation is the integral characteristic reflecting the pollution of the surface atmospheric layer through which the precipitation passes. The dynamics of change of this characteristic throws light on the effect of the general atmospheric circulation on the dispersion of green-house gases and aerosols over a certain territory. It is interesting to investigate the content of hydrocarbonates in precipitation, as the indicator of atmospheric pollution by carbon oxide (a green-house gas). Table 2 shows the relationship between HCO3- content and the ɪɇ of the liquid-phase media.

Table 2. Species

Speciation of carbonic acid in aqueous solutions of various ɪɇ (molar %). ɪɇ 4

5

6

7

8

9

10

11

[H2CO3 ]

99.7

97.0

76.7

24.9

3.22

0.32

0.02



[HCO3- ]

0.3

3.0

23.3

74.98

96.7

98.84

71.43

20.0

[CO32- ]







0.03

0.08

3.84

28.55

80.0

As part of the studies we have determined the dynamics of the change in the chemical composition of precipitation measured at some stations (Tashkent – Observatory, Kaunchi), by applying the commonly used indices (ɪɇ, electric conductivity, total ions, SO42-, NO3-, HCO3-, Cl-, Ca2+, Mg2+, NH4+, Na+, K+). Appropriate trends were revealed.

Problems of Air Quality in Tashkent City

Figure 11.

389

Correlation between the water-soluble components in precipitation.

Possible relationships between the levels of precipitation pollution by green-house gases species and aerosols (HCO3-, NO3-, Ca2+, Mg2+) and by masses of emissions over the investigated territory have been studied. After analyzing the data on the chemical composition of the atmospheric precipitation measured over the territory of Tashkent city and Tashkent province, it was found that the local sources of atmospheric pollution have a certain effect on the chemical composition of precipitation (Figure 11). Simulation of the Effect of Atmospheric Phytotoxic Trace Components on Land Vegetation The study of the effect of individual atmospheric pollutants on plants under simulated fumigation conditions in chambers is a compulsory element for the assessment of the effect of atmosphere pollution on land vegetation. The concentrations at which atmospheric pollutants cause irreversible changes in the vegetal cell after 30 minutes and 4 hours of exposure were estimated experimentally. Fresh leaves of laboratory plants were used to study the relationships between the concentration of individual pollutants in the fumigation chamber and the peroxide activity, the intensity of peroxidation of lipids and the content of photo-synthetic pigments in the vegetal cell.

G. A. Tolkacheva

390

Based on the output of the conducted experiments, regional ecological standards for the quality of the atmosphere, with respect to the major air pollutants were developed, taking into account the risk for the land vegetation (trees and agricultural crops and plants) (Ɍɚble 3). Maximum admissible concentrations (MAC) of major atmospheric pollutants for humans and for vegetation.

Table 3.

One-time MAC for humans mg/m3 0.5 0.16 0.2 0.02 0.085

Pollutant SO2 O3 NH3 HF NO2

One-time MAC for vegetation mg/m3 0.3 0.061 0.091 0.0096 0.04

Average daily MAC for humans mg/m3 0.05 0.03 0.04 0.005 0.04

Average daily MAC for vegetation mg/m3 0.029 0.029 0.04 0.0033 0.015

As was noted before, the index of atmosphere pollution (IAP) is used in the overall state system of atmosphere monitoring and evaluation of air quality, for the estimation of the degree of air pollution. MACs for ecology were determined for the evaluation of atmospheric air quality in industrial agglomerations, which take into account the harmfulness of pollutants for land vegetation. Thus, new IAPs were developed for vegetation which differ from IAPs estimated on the basis of sanitary-and-hygienic MACs. Examples are shown in Figures 12, 13, and 14.

0,4 0,35 0,3 0,25

for plant

0,2 0,15 0,1 0,05 0 I

Figure 12.

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

Dynamics of the index of atmospheric pollution by SO2, Tashkent, 2001.

Problems of Air Quality in Tashkent City

391

7 6 5 4 3 2 1 0 I

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

Dynamics of the index of atmospheric pollution by NO2, Tashkent, 2001.

Figure 13. 3,5 3 2,5 2 1,5 1 0,5 0 I

Figure 14.

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

Dynamics of the index of atmospheric pollution by NO2, Angren, 2001.

In our opinion, it is evident that the indices of the atmospheric pollution calculated with the use of ecological standards of quality, present more fairly the impact of individual pollutants on the ecosystem. The conducted experimental studies let us propose four levels characterizing the degree of atmospheric pollution by a given pollutant (i): admissible (Ji ” 1)

the pollutant concentration is not harmful for vegetation

increased (1 <Ji < 3) substantial (3 < Ji ” 5) high (Ji > 5)

the pollutant concentration is dangerous for vegetation

An IAP is usually estimated by mean-annual concentrations of harmful trace components, as follows: cj

§ q · ¦ ¨¨ i ¸¸ . i 1 © MPC ¹ m

Jm

G. A. Tolkacheva

392

7,87

Angren

8,98 7,72

Almalyk

10,75 8,67

Tashkent

7,91

0

2

4

2001

Figure 15.

6

8

(J) 10

12

2002

Integral atmospheric pollution (for 5 main pollutants) in several Uzbekistan towns, in 2001 and 2002.

Here, it is recommended to calculate the partial index values for each pollutant, and then to calculate the total index as a sum of the five highest values. Thus, we estimated the integral pollution of the atmospheric air as a sum of IAPs of each pollutant in a given year for five main atmosphere pollutants posing risk for vegetation. Mean annual indices of integral pollution of the atmospheric air were calculated for the big industrial centres of Uzbekistan. Some of them are presented on Figure 15. Following the specification for individual pollutants, four levels characterizing the degree of the integral atmospheric pollution can be considered for five main pollutants: admissible (J ” 5)

the concentration of the pollutants is harmful for vegetation

increased (5 < J < 8) substantial (8 < J ” 11)

the concentrations of the pollutants are dangerous for vegetation

high (J > 11)

Figure 15 shows that in all investigated towns of Uzbekistan an increased (5 < J < 8) or substantial (8 < J < 11) level of atmospheric pollution was observed in 2001 and 2002. Conclusion Analysis of the current state of the pollution of the atmospheric air in Tashkent city and of the system of the air quality monitoring has revealed the following: ¾ It is necessary to develop a united information system consisting of automated stations for air quality control; ¾ There is not any system of measuring the dust particles, which is important for the population health; ¾ It is necessary to develop the models of pollutant transfer and transformation for the conditions of the urban environment, taking into account the meteorological parameters. Acknowledgment I thank Mr. A.A. Azizov, Chief of Ecological Program of the State Committee on Science and Technology of the Republic of Uzbekistan, and Ms. N.G. Verechagina, Chief of the Laboratory of NIGMI for financial support of this work. I thank L. Shardakova, Y. Kovalevskaya, T. Smirnova, N. Rahmatova, L. Usmanova and N. Akinshina for their assistance with carrying out this work.

Precipitation Quality in Different Zones of the Tashkent Region in Relation to Photo-Chemical Reactions Tatyana Smirnova and Galina Tolkacheva Hydrometeorological Research Institute Uzhydrome of the Republic of Uzbekistan, 700052 Tashkent, K. Makhsymov, 72 Key Words: Photochemical reactions, Oxidative capacity, Local sources, Mineralization, Underlying surface

Introduction Since many years there has been increased interest in the investigation of chemical processes taking place in the polluted atmosphere. Definite relationships exist between chemical reactions which occur in the atmosphere in the gas, liquid and solid phases under stationary conditions. It is interesting to study comprehensively the relationships and the nature of the processes taking place in a liquid phase and in aerosols, as well as the role of chemical atmospheric processes in the formation of secondary aerosols. The increase of the secondary aerosol concentration in the atmosphere causes a decrease in visibility and the formation of acid rains which have a negative effect on human health, vegetation, soil, buildings and constructions (Ravinskii and Egorov, 1990). Tashkent province was selected for the investigations. This area is known for its typically extreme continental climate, abundance of days with sunshine during the year which means a high solar radiation, high mean-annual air temperature and complex orography. Large industrial enterprises are located in this area. The main emissions of these plants into the atmosphere are: dust (with an abundance of heavy metals), oxides of nitrogen, sulphur and carbon, and other compounds of nitrogen, sulphur and fluorine. It is known that pollutants fall from the atmosphere to the ground surface due to various processes including the dry absorption of gases, humid fallout with precipitation and gravitational deposition of suspended solid particles. The processes which influence the chemical composition of atmospheric precipitation can be investigated using the chemical reactions taking place in atmosphere. It is interesting to follow the possible ways of formation of different compounds in the atmosphere in the selected area in relation to the composition of the emissions of harmful substances into the atmosphere. Objects and Methods of Investigations In this section we describe the methodology used to study the effect of emissions from the local sources of atmospheric pollution on the processes of secondary aerosol formation. The basic admixtures selected as the main objects of investigations were: dust and oxides of sulphur and nitrogen which are the main components of the emissions from the industrial plants in Tashkent province. 393 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 393–402. © 2006 Springer. Printed in the Netherlands.

T. Smirnova and G. Tolkacheva

394

In the first stage of the studies an assessment of the dynamics of the main emissions from the industrial plants and the contents of each component was performed, based on the data from statistical reports (Glavgidromet, 2001). At the same time, the changes of the chemical composition of the gas phase, atmospheric aerosols, and precipitation during the same period were analyzed, and possible impacts of atmospheric fallouts on the environment were assessed. The effect of emissions on the chemical composition of atmospheric precipitation was investigated taking Almalyk town as an example. A lot of industrial objects are located in Almalyk town: construction enterprises and chemical plants; Almalyk mining and smelting plant. Based on the methodical guideliness formulated in the Main Geophysical Observatory and SANIGMI (Gidrometeoizdat, 1991; SANIGMI, 1993a), sampling of precipitation was conducted for 14 years. Sampling and analysis of sulphur and oxygen oxides in atmosphere were performed using a standard technique (Andreani-Aksogoglu, 1996; SANIGMI, 1993b). The chemical composition of atmospheric precipitation was analyzed by the methods of ion chromatography, atomic adsorption, photo-colorimetry and other techniques, in accordance with documented methodical guidelines (Brimblecombe, 1988; SANIGMI, 1993b; Gidrometeoizdat, 1996; Isidorov, 2001). Outcomes and Their Consideration As was mentioned above, at the first stage of the investigations we assessed the dynamics of the emissions of sulphur and nitrogen dioxides and of dust which pollute the atmosphere in Almalyk, the levels of atmospheric pollution of these components and their content in humid and dry atmospheric fallouts. The town of Almalyk was chosen as a model for the investigation of the chemical reactions in the atmosphere causing the formation of nitrates and sulphates in precipitation. Total annual emissions and annual mean concentrations of each component were also calculated. The results obtained are presented in Table 1 (for 2 years).

Table 1.

Emissions and concentrations of atmospheric pollutants in air, precipitation and dry atmospheric fallouts (DAF).

Pollutant Sulphur dioxide Nitrogen oxides Nitrates (water-soluble) Sulphates (water-soluble) Suspended substances Lead , t/y Arsenic, t/y

Emissions into atmosphere 1000 t/year 1999 88.9 1.02

2000 87.7 1.01

Mean concentration in air mg/m3 1999 2000 0.04 0.04 0.05 0.05

0.04

0.03

3.81

3.8

0.2

0.2

1.93 17.6

2.64 33.0

0.12

0.12

in precipitation mg/dm3 1999 2000

DAF mg/dm3 1999 2000

8.09

7.99

5.81

0.913

41.27

34.51

49.71

14.06

Precipitation Quality in Different Zones of Tashkent Region

395

After calculations using the experimental method it was found that in 1999 in Almalyk the total SO42- deposition with liquid precipitation was 151.12 kg/ha/y (monthly average – 12.6 kg/ha) and the total SO42- deposition with dry fallout (DAF) was 139.2 kg/ha/y, (monthly average – 11.6 kg/ha). Figure 1 presents the dynamics of the changes in atmospheric pollution by sulphur and nitrogen oxides and dust.

Figure 1.

Changes in the average levels of atmosphere pollution in Almalyk town.

The industrial enterprises of Almalyk, heating and production of electric energy are the main sources of sulphur dioxide emission into the atmosphere (90% of all emissions). The main source of emission of nitrogen oxides is traffic. During the last years a small increase of sulphur dioxide and dust concentrations in atmosphere was recorded. Correlation factors between the pollutant emissions into the atmosphere and their concentration in atmospheric precipitation have been calculated. Table 2 lists the values obtained. The correlation factors indicate the presence of a certain relationship between the pollutant emissions into the atmosphere and their concentration in atmospheric precipitation.

Table 2.

Correlation factors between pollutant emissions into the atmosphere and concentrations of these substances in atmospheric precipitation.

Pollutants in the atmosphere

Dust (1000 t)

Pollutants in atmospheric precipitation Ion sum mg/dm3

Correlation factor - 0.485

SO2 (1000 t)

SO4 - mg/dm3

0.56

3

0.64

NOx (1000 t)

-

NO3 mg/dm

Nitrogen and simultaneously formed non-organic (ɇ2Ɉ2) and organic (ROOH) peroxides and peroxyacylnitrates are the strongest photo- and phyto oxidants. The increase of the concentration of these secondary pollutants in the atmosphere causes a decrease in bio-

T. Smirnova and G. Tolkacheva

396

productivity and the degradation of a natural vegetation cover over vast territories (Mazin and Krichian, 1988) and, consequently, has a negative effect on the assimilation of one of the main green-house gases, carbon dioxide. The positive trend in the troposphere ozone concentrations is most likely determined by the increase of emissions of the ozone precursors, nitrogen oxides and volatile organic compounds (VOC) from anthropogenic sources. Photo-chemical reactions taking place during the action of intensive solar radiation, high temperature and humidity on the air mass containing pollutants are the main source of the formation of ozone in atmosphere. The formation of ozone in the polluted atmosphere proceeds as follows: under the effect of radiation (wavelength range 310 - 1180 nm), ozone is decomposed with the formation of atomic oxygen in the ground state (Ɉ(3P)). O3+hȞ o Ɉ(3P)+Ɉ2 The oxygen atoms recombine with oxygen to ozone: Ɉ(3P)+O2+Ɇ o Ɉ3+Ɇ* where Ɇ is a molecule of nitrogen or oxygen which obtains excessive vibrational energy from the intermediate complex (Ɉ2 … Ɉ). During the absorption of quanta by ozone in the nearest UV band in the 290 - 310 nm range, the following reaction takes place in the atmosphere: O3+hȞ o Ɉ(1D)+Ɉ2(1'd) The major part of the unstable oxygen being formed in the troposphere is deactivated through collision with molecules of nitrogen or oxygen: Ɉ(1D)+Ɇ o Ɉ(3P)+Ɇ* Ɉ+Ɉ2+Ɇ o Ɉ3+Ɇ A fraction of the unstable oxygen reacts with water vapour to produce hydroxyl radicals: Ɉ(1D)+ɇ2Ɉ o 2ɇɈ+120.5 kJ (Shreier, 1986) Hydroxyl radicals are very reactive, and take part in reactions with many pollutants to make the atmosphere “cleaner”. It should be taken into account that the concentrations of trace gases in the polluted atmosphere can be much higher than in the clean one, so the rates of second order reactions increase significantly. It is most likely that ozone formation in the polluted atmosphere takes place during the oxidation of nitrogen oxides to nitrogen dioxide. Under certain radiation conditions, a pseudobalance between nitrogen oxide and dioxide is maintained. In such cases, the increase of the ozone level in the atmosphere is the evidence for the decrease of the nitrogen oxide concentration therein. Consideration of other reactions of organic components possible in the polluted atmosphere confirms the statement that the main source of ozone in the troposphere is the photolysis of nitrogen oxides: NO2+hȞ (Ȝ<310 nm) o Ɉ(3P)+ NO Ɉ(3P)+O2+Ɇ o Ɉ3+ Ɇ Ɉ3+NO o Ɉ2+NO2

Precipitation Quality in Different Zones of Tashkent Region

397

Thus, the considered mechanisms of ozone formation in the surface layer are mostly characteristic for urban territories. Specific nature and climatic conditions of the Tashkent province, the abundance of the sunny days and various conditions of natural zones determine, to a certain extent, the nonhomogeneous conditions necessary for the photo-chemical reactions in the atmosphere. On the other hand, the concentration of large industrial enterprises on the territories of river valleys creates definite conditions for the accumulation of polluting components along the mountain ranges. The industrial towns of Tashkent province are stretched along the valleys of the Akhangaran, Chirchik and Syrdarja rivers. These territories are characterized by a higher climatic potential of pollution, by a daily mountain and valley circulation which promotes the high level of atmospheric pollution over the territory and the formation of photo-chemical smog. The pollutants emitted into the atmosphere are later transformed to sulphates and nitrates on the dust particles, under conditions of increased humidity and solar radiation. These transformations can be described by a cyclic process incorporating the following photochemical reactions: CO + OH (+O2) o CO2 + HO2 HO2 + O3 o OH + 2 O2 CO + O3 o CO2 + O2 SO2 + OH- + hȞ o HSO3HSO3- + O2 o HO2 + SO3 SO3 + H2O o H2SO4 . O3 + NO2 o NO3 + O2 . NO3 + hȞ o NO + O2 . . NO3 + organics o HNO3 + R An analysis of the systematic data on the change of the surface ozone concentration over the towns of Tashkent province has been performed. Figure 2 presents the mean longterm values of the ozone concentrations in Almalyk, Tashkent and Chirchik towns. It was found that during the whole period of the observations, the ozone concentrations (mean long-term values) generally exceed the values of admissible concentrations. In Chirchik town, even the minimum values of ozone concentrations are rather high. The lowest values of ozone concentrations were recorded in Almalyk. During the last years in Tashkent the ozone concentration has decreased a little. The mean long-term values for the towns of Tashkent province are comparable with the values of ozone concentration in a surface layer for Central Europe territory (0.05 - 0.09 mg/m3).

398

Figure 2.

T. Smirnova and G. Tolkacheva

Average, maximal and minimal long-term values of ozone concentrations in the surface layer, recorded in the towns of Tashkent province in 1995-2000.

On the other hand, during the last years the problems of anthropogenic impacts on nature (and especially, the pollution) sparked a great interest in the problem of transfer and deposition of various chemical substances over long distances. The next stage of the work included the study of the dynamics of the chemical composition of precipitation sampled at the meteostations of Almalyk, Tashkent and Chirchik, with the purpose of estimating the possible impact of the local pollution of the atmosphere on the composition of the precipitation. As already discussed, we can identify the major substances which pollute the atmosphere: sulphur dioxide and nitrogen oxides, and products of their transformation. During the interaction with water droplets and falling rain, the acids and their salts (even more toxic) are formed from these gases. Acid rain leads to considerable negative ecological effects. Due to the presence of free oxygen, the atmosphere is a system with oxidizing abilities, so almost all reactions of sulphur and nitrogen compounds lead to the formation of sulphates and nitrates, as the highest forms of sulphur and nitrogen oxidation. Gas-phase oxidation can utilise oxygen, nitrogen, hydrogen peroxide as molecular oxidants. However, it is more effective with the free radicals involved. Photo-chemical processes play an important role in the formation of radicals (Izrael et al., 1983). It was found in Almalyk that the concentration of sulphates in atmospheric precipitation is 4 times higher than that of nitrates. This is caused by higher emissions of sulphur oxides in Almalyk than in Tashkent or Chirchik. As was mentioned before, the ozone concentration in Almalyk is the lowest (Figure 2). Probably, during the oxidation of sulphur in the atmosphere up to S(VI), ozone partially participates in the chemical transformations. In order to identify the role of ozone in the transformation process of nitrogen and sulphur oxides into sulphates and nitrates, we used long-term data to construct seasonal graphs of the dynamics of change of the ozone concentrations by month, as well as seasonal graphs for sulphates and nitrates.

Precipitation Quality in Different Zones of Tashkent Region

Figure 3.

399

Dynamics of the change of the mean long-term ozone concentrations by month (1995 – 2000) recorded in the towns of Tashkent province.

The increase of ozone concentrations during summer months (Figure 3) in all towns confirms the increase of the rates of photo-chemical reactions in the atmosphere followed by ozone formation. Possibly, there is an interrelation between ozone content in the atmosphere and the concentrations of pollutants in the atmospheric precipitation. It was found that in the course of reactions taking place in a polluted atmosphere of Almalyk and Tashkent towns, the maximum sulphate concentrations in the atmospheric precipitation are recorded in August and in Chirchik in September. In Tashkent, the maximum sulphate concentration in precipitation is observed also in August while in Chirchik it is in June and August. In Almalyk, the fluctuations of nitrate concentration are rather low (Figure 4). The maximum concentrations of sulphates and nitrates in precipitation during the considered period are related to a preceding long dry period. During that time the accumulation of component emissions takes place under the conditions of high dust pollution determined by climate conditions and underlying surface as well as intensive solar radiation; the reactions of transformation of sulphur and nitrogen to sulphates and nitrates on aerosol particles. As a result of the conducted studies the pattern of possible transformations and migration of sulphur and nitrogen oxides within the atmosphere of the investigated region taken as the experimental natural test-field is proposed (Figure 5). Conclusions When studying the chemical composition of atmospheric precipitation it is possible to assess the pollution of the atmosphere, and consequently, to evaluate the ecological conditions of the region. For the evaluation of the present state of the atmosphere for the exemplary cases of Almalyk, Tashkent and Chirchik towns, the composition of emissions, using statistic reports, as well as seasonal dynamics of emissions has been investigated. The data from systematic observations of the atmospheric pollution in these towns (data of observations made at three stationary posts) were generalized.

T. Smirnova and G. Tolkacheva

400

Under the conditions of the arid climate and anthropogenic pollution, a negative effect of atmospheric precipitation on the ecology and public health in the investigated region was detected, due to the washing out of pollutants from the atmosphere with atmospheric precipitation which falls on environmental objects and results in the deterioration of the ecological conditions of the region.

80

mg/dm3

NO3

Almalyk

S04

70 60 50 40 30 20 10 0 April

20

mg/dm3

May

NO3

June

July

August

September

S04

Oktober

Tashkent

15 10 5 0 April

May

m g/dm 3

NO3

June

July

August

September

S04

Oktober

Chirchik

16 14 12 10 8 6 4 2 0 April

Figure 4.

May

June

July

August

September

Oktober

Dynamics of the changes of the sulphate and nitrate concentrations (mean longterm values for 1995 – 2000) in atmospheric precipitation in the towns of Tashkent province.

Precipitation Quality in Different Zones of Tashkent Region

401

EMISSIONS INTO THE A T M O S P H E R E

SULPHUR OXIDES

NITRIC OXIDES

CHEMICAL COMPOSITION OF THE ATMOSPHERE

Enriched aerosols Fe, Cu, Pb, Zn, CuO, NH3, O2, CARBON SO2 aqueous phase (hydrosols)

NO, NO2, N2O4 aqueous phase (hydrosols) O3, ɇɈ-, ɇ+, H2O2

NOx(NO + NO2) gas phase

WASHING

SO2 gas phase

OUT WITH PRECIPITATION, DRY DEPOSITION

REACTION PRODUCTS

NITRATES NO3

Figure 5.

SULPHATES SO4

Pattern of possible transformations and migration of sulphur and nitrogen oxides in the atmosphere of the region investigated.

402

T. Smirnova and G. Tolkacheva

References Andreani-Aksogoglu, S.; Impact of air pollutants on ecosystems; a review, Paul Schrrer Inst. Berichte 96-05 (1996) Abteilung Luft Fremdstoffe. Brimblecombe, P.; Chemistry and composition of atmosphere. Ɇir, (1988) (in Russian). Gidrometeoizdat; RD 52.04.186-89 Guiding document. Guidelines on the control of atmospheric pollution – Ɇ. Gidrometeoizdat, (1991) (in Russian). Gidrometeoizdat; Manual on chemical composition of a surface water. – L.: Gidrometeoizdat, (1996) (in Russian). Glavgidromet; Review of atmospheric pollution and emissions of harmful substances on the territory serviced by Glavgidromet of Republic of Uzbekistan for 2000. Tashkent (2001) (in Russian). Isidorov, V.; Organic chemistry of atmosphere. Khimizdat, Sankt-Petersburg (2001) (in Russian). Izrael, Yu., I. Nazarov, A. Pressman, F. Yarovinskii, Ⱥ. Ryaboshapko and L. Filipov; Acid rains. Gidrometeoizdat, (1983) (in Russian). Mazin, I. and Ⱥ. Khrichian (eds); Clouds and cloud atmosphere: Reference book. Gidrometeoizdat, (1989) 44-72 (in Russian). Ravinskii, F. and V. Egorov (eds); Acid fallouts. Long-term trends – L., Gidrometeoizdat, (1990) 119-120 (in Russian) SANIGMI; RD 4600430.21 Method guidelines. Method of measurement of mass concentration of fluorine and chlorine anions, hydro-phosphates, nitrates and sulphates in atmospheric aerosols, using the technique of ionic chromatography. SANIGMI (1993a) (in Russian). SANIGMI; RD Uz 52.25.23-96 Method instructions. Method of measurement of mass concentration of anions of fluorine and chlorine anions, hydro-carbonates, nitrates and sulphates in atmospheric precipitation and surface water by the method of ionic chromatography – Ɍ. SANIGMI (1993b) (in Russian). Shraier, D. (ed); Heterogeneous chemistry of atmosphere. Gidrometeoizdat, (1986) 44-72 (in Russian).

Influence of Atmospheric Aerosol Contamination on the Regional Climate in Central Asia Boris B. Chen and Valery M. Lelevkin Kyrgyz-Russian Slavic University, Bishkek, Kyrgyzstan Key Words: Lidar, Aerosol, Backscattering ratio, Backscattering coefficient, Extinction, Backtrajectory

Introduction At present, in the mountain regions of Tien-Shan and Pamir reduction and vanishing of glaciers is occurring. According to glaciological data, 1081 glaciers disappeared at Pamir and Alai between 1957 and 1990, and 71 glaciers have completely thawed at Zailijsky Ala Tau. The glaciers recede with an average velocity of 8 meters per year. Not only is the linear size of glaciers reduced, but also their volume. This influences significantly river waters, vegetation and Central Asian climate. There are several reasons for glacier destruction, one of which is the global rise of temperature. According to the World Meteorological Organization experts, during the last decades the increase in temperature of the ground air layer was 0.5ɨ, and in the next 30-50 years it could cause a significant rise in global temperature. The other reason is connected with the drying of the Aral Sea and the associated intensified wind erosion of the dried bottom. In a dust cloud, the suspended aerosol particles with admixtures of agricultural fertilizers and other components of industrial and home wastes predominate. The dust, lifted into the air, reaches high heights and is transported over significant distances. Dust accumulation on the surfaces of the glaciers leads to their pollution and causes intensive ice thawing. Scientists of St.-Petersburg University tested this hypothesis experimentally, on the basis of the analysis of aerosol samples taken from different glaciers of Pamir and Tien-Shan, near the Aral Sea. The obtained data, however, did not allow them to prove unambiguously the strong influence of Aral Sea drying on dust accumulation on glacier surfaces, though a constant increase of the concentration of NaCl and Ba compounds, characteristic for the seas, has been observed in ice sediments. The glacier pollutions with dust from weathering belong to natural drivers of thawing. The dust comes from Afghanistan, China and desert regions surrounding Tien-Shan. Much of it is brought by dust storms originating in the deserts of Central Asia. The existing opinion, that the main reason of climate change is the increased accumulation of greenhouse gases in the atmosphere, is not convincing enough. Recent investigations have shown that the stratosphere aerosol (SA) is one of the basic factors influencing the radiative balance and climate (Chen and Lelevkin, 2000). SA influences systems with high levels of exiting long-wavelength radiation, for example systems with cloudless skies and warm spreading surfaces. Cooling at the expense of aerosol (the albedo effect in the short-wave region of a spectrum) predominates over heating (the greenhouse infra-red effect) for all regions, excluding the polar regions in winter (Chen and Lelevkin, 2000). In the background phases, the heating effect is observed in the stratosphere during the cold half-year, and the albedo effect is observed in the troposphere. During the warm halfyear, predominate cooling of the stratosphere and heating of the troposphere is observed. Estimations of temperature change at the expense of background SA shows a drop of ground 403 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 403–414. © 2006 Springer. Printed in the Netherlands.

B. B. Chen and V. M. Lelevkin

404

temperature in northern regions by 0.16-0.18 oɋ per year, on the average. In high mountain regions the drop is 0.11-0.13 oɋ, which compares well to model calculations predicting that the background SA reduces the average global temperature of the ground layer by 0.1 oɋ. Up to 20 grams/m2 of dust accumulate on glaciers per year. Natural phenomena such as eruptions of volcanoes, earthquakes, mud flows, dust storms, floods, droughts, as well as anthropogenic catastrophes – nuclear tests, fires at petroleum slits, wood cutting and drying of the Aral Sea – exert a huge influence on the pollution of the atmosphere. Methodology In Central Asia, the measurements of vertical profiles of concentration and the optical characteristics of stratospheric and tropospheric aerosol have been conducted since 1988 at the Lidar Station Teplokluchenka (Kyrgyz Republic), with the use of a laser-location method. The high mountain Lidar Station Teplokluchenka (LST) is located over 2000 m above sea level, southeast of a high mountain lake Issyk-Kul in Central Tien-Shan (42.5o N, 78.4o ȿ). It is a unique scientific station for complex monitoring of tropospheric and stratospheric aerosol by lidar in the center of the Asian part of the global geoecological system. LST was established in 1987. Unique series of observations of the optical characteristics of stratospheric aerosol were obtained at the station in 1988-2001, with the lidar «Ɇɚkɟt-1». Since 2002 tropospheric and stratospheric aerosol monitoring has been conducted with a modernized multi-wavelength lidar. The lidar can operate in the two following modes: x Simultaneous registration of the intensity of backscattering radiation at three laser wavelengths (355, 532 and 1064 nm), and of the Raman backscattering by atmospheric nitrogen (387 nm), in analog or photon counting modes of operation; x Registration of two perpendicular polarization components of the backscattering signal at the 532 nm wavelength with the use of Vollaston prism. Experimental control and data processing are performed with an IBM PC. The general parameters of the modernized lidar are presented in Table 1.

Table 1.

Specifications of a modernized multiwavelength lidar.

Transmitter Lotis LS-2137 Laser Wavelength 1064, 532, 355 nm 300, 120, 80 mJ Energy Angular divergence 0.6 min Pulse duration 15-20 ns 10 Hz Repetition rate

Receiver 50 cm Telescope diameter 1.5-15 min Field of view Mechanical chopper Photomultipliers PMT-100, PMT-83 25MHz / 12bit ADC 200 MHz Photon counter

The methodology of multi-wavelength lidar sensing of the atmospheric aerosol, the processing of the backscattering signal and the reception of optical and microphysical characteristics of the aerosol are given in Chen et al. (2002, 2004).

Influence of Atmospheric Aerosol Contamination on the Regional Climate

405

The properties of aerosol measured at LST are: x Aerosol optical properties: scattering, absorption, aerosol optical depth; x Aerosol physical properties: mass, distribution spectra of aerosol particle size, Lidar ratio. x Vertical distribution of aerosols: extinction coefficient, backscattering coefficient, single scattering albedo. The following integral parameters of the aerosol spectrum are also calculated: characteristic radius of particles, Pm; particle surface area per unit of air volume, Pm2/cm3; particle volume per unit of air volume, Pm3/cm3. Influence of natural catastrophes The eruption of Pinatubo volcano (Philippines, June 15-16, 1991) was one of the most powerful eruptions in the last century. Huge amounts of gaseous and aerosol substances were thrown into the atmosphere (Figure 1). Evaluations based on satellite observations showed that the expelled mass of sulphur dioxide was approximately 20 million tons (Bluth et al., 1992), and rendered a powerful influence on radiating processes in the atmosphere as well as on the transformation of the ozone layer. Oxidation of sulphur dioxide leads to formation of a fine dispersion of sulphuric acid aerosol. Figure 1 shows that the concentration of SA increased sharply for 3 months, reaching a maximal value in January 1992, and than decreased until May. The increase of the optical thickness in the initial phase is explained by the fast growth of particles of poorly absorbing sulphuric acid aerosol.

Figure 1.

Monitoring of integral backscattering coefficient in the height range 15-30 km.

Influence of stratospheric aerosol on total amount of atmospheric ozone Volcanic eruptions can influence the total amount of ozone in the atmosphere (total ozone). The mechanisms of this influence can be very different. For instance, in non-volcanic periods the accumulation of ozone may take place, due to the presence of sulphur dioxide (SO2, the sulphurous gas) in the atmosphere and its photooxidation by air oxygen (Ivlev et. al., 1990). We shall examine some mechanisms of the accumulation and exhaustions of the

406

B. B. Chen and V. M. Lelevkin

stratospheric ozone under the influence of sulphuric-acid aerosol for different conditions of the atmosphere. The background period The analysis of the heights of the location of the backscattering ratio maximum Rmax over the Central Asian region (with an average height of 18.25 km, 17.62 km in the cool and 19.0 km in the warm half-year) indicates that the maximum concentration of the background stratospheric aerosol (SA) is observed mainly in the stratosphere, in all seasons of the year (Chen and Lelevkin, 2000). Consequently, in the background periods the sulphurous gas SO2 arriving from the troposphere is not a direct precursor of the aerosol. Photodissociation of cɚrbɨnyl sulphide and subsequent formation of the sulphurous gas in the lower stratosphere seems to play the main role in determining the background SA. The results of the experiment (Chen and Lelevkin, 2000) show, that before the Pinatubo volcano eruption, in the background period the area of 2429 km was marked out with local minimums in the correlation function at which the main mass of the background aerosol was concentrated. This happened at the height of the maximum ozone concentration (2427 km), not in the field of the Junge aerosol layer, so one can expect that the ozone accumulation takes place there, i.e. O3 is generated from the photooxidation of SO2 by air oxygen (Ivlev et. al., 1990): 3 SO2 ( 3B1 )  O2 o SO3  O( 3P ) . In winter (December), a zonal (western) circulation prevails in the stratosphere at moderate latitudes. In January and February, the circulation in the stratosphere is unstable, and the meridianal transference prevails over the zonal one (Stolypina, 1981). This is, because the circumpolar cyclonic whirlwind shifts to the south, the Pacific maximum shifts to the north, and both of them become immobile. An increase in the height of SA maximum location and in the maximum optical thickness is observed (Chen and Lelevkin, 2000). Volcanic sulphur dioxide and ozone As a result of the Pinatubo volcano eruption in the Philippines in June 1991, great amounts of gaseous and aerosol matter were thrown into the atmosphere. The condition of the SA and variations in its parameters greatly influenced both the radiative processes and the ozone layer transformation. In summer the SA transformation at altitudes lower than 20 km is caused by the transfer of western air masses. In the higher stratosphere and at eastern circulation, there are no conditions promoting any aerosol transfer from tropical latitudes into moderate zones of the northern hemisphere. In summer the meridianal circulation in the stratosphere becomes practically zero (Stolypina, 1981). The aerosol was registered at the heights larger than 20 km in late October of 1991, under western circulation establishing in a moderate zone of the stratosphere. For all of this, the summer stratospheric anti-cyclone was destructed and new favourable conditions occurred in the upper stratosphere for the aerosol transfer from the tropical zone into the moderate zones of the northern hemisphere. During this period one observed an increased SA maximum (Chen and Lelevkin, 2000). From June 1992 to January 1993, the SA increased ten times compared to the background SA before the volcano eruption (Chen and Lelevkin, 2000). Then, the concentration of SA gradually decreased back to the level of 1988-1989. The change was explained by the fact that at volcanic eruptions of the explosive type not only sulphate

Influence of Atmospheric Aerosol Contamination on the Regional Climate

407

particles of different sizes appeared in stratosphere, but also a great amount of sulphur dioxide (Turko et al., 1983). The thickness of the SA increased after sulphur dioxide reached the stratosphere and was oxidised to sulphuric acid vapours, which condensed together with water vapour on the particles already present in stratosphere, or formed new particles by homogenous nucleation from the gas phase. In our case study, these processes evidently continued up to the end of 1992 and to the beginning of 1993. The measurements showed that major masses of volcanic aerosol in the first period after the volcano eruption were located in layers of 16-18 and 23-25 km (Chen and Lelevkin, 2000). During the period following the sedimentation of particles, the formation of aerosol from sulphur dioxide in the layer of maximum stratospheric ozone concentration (26-28 km) occurred. Therefore, the reduction of total ozone occurred (Figure 2), due to the photooxidation reaction of SO2 (Toktomyshev and Semenov, 2001): 3

SO2 ( 3B1 )  O3 o SO3  O2 .

From June 1992 until February 1993, a sharp deceleration of the aerosol concentration depletion was observed. As a result of the reaction SO3  H 2 O o H 2 SO4 . stratospheric aerosol was formed during this period. During March and August 1993, the SA concentration diminished nearly two times and the total ozone sharply proliferated (Figure 2) as a result of ozone generation from SO2 photooxidation by air oxygen (Okabe, 1981): 3

Figure 2 .

SO 2 ( 3B1 )  O2 o SO 3  O ( 3P )

Joint distribution of Rmax and total ozone (1 – total ozone, 2 – Rmax).

Empirical links between SA and general concentration of ozone The analysis of empirical links between the SA and the general concentration of ozone (GCO) has shown, that during all the active period of the Pinatubo volcano eruption, the coefficient of the linear correlation between SA and GCO was equal to r=0.87r0.07, while the reliability of the linear correlation was P=0.99.

408

B. B. Chen and V. M. Lelevkin

During the background period before the volcano eruption, the SA was formed in the stratosphere due to the photo-oxidation of stratospheric SO2, and the coefficient of correlation between the SA and GCO was negative (r=-0.46r0.17, with P=0.95). When products of the volcano eruption came to our latitudes, O3 was absorbed in the sulphate aerosol during formation of SA from SO2 (from June 1993 to February 1993), so the concentration of O3 was reduced. Thus the negative correlation between SA and GCO increased comparing to the background period: r=-0.76r0.12, P=0.99. During ozone generation from the photo-oxidation of sulphur dioxide by air oxygen, which occurred between March 1993 and August 1993, the coefficient of correlation became positive again: r=0.88r0.07, P=0.99. Thereby, the process of SA relaxation after the volcanic eruptions influenced the stratospheric ozone content in two ways: by reducing the general concentration of ozone through SO2 photo-oxidation, and by increasing GCO as a result of ozone generation from SO2 photo-oxidation by oxygen of the air. Hence, besides the known reasons of so-called “local ozone holes” (ozone concentration reduction) over Central Asia (Toktomyshev and Semenov, 2001), there are other mechanisms of ozone reduction in the stratosphere connected with the ozone absorption by the stratospheric sulphate aerosol, and with sulphur dioxide photo-oxidation during the background period. Military actions in Iraq and atmosphere pollution At present, the problem of anthropogenic pollution of the environment increases in scale and gradually becomes a political, economical and scientific problem, calling for the global public concern. Such deliberate human actions as war contribute to atmosphere pollution, when huge amounts of dust, smut and carcinogenic materials originating from fires are thrown into the atmosphere as a result of military actions. During February and before the military actions in Iraq, no significant pollution of the lower troposphere coming from the Middle East was recorded. An aerosol cloud consisting of dust (sand) particles, soot and smoke coming with south-west air flows was registered by the Teplokluchenka station on March 20. The day before, a heavy sand storm occurred in the region of military actions. Maximum backscattering ratio R higher than 100 was observed at the wavelength of 1064 nm (Chen et al., 2004). The size distributions of the particle concentrations show, that a layer 1.0 km thick (2.5-3.5 km) stands out between the layers of 1.5-2.5 km and 3.5-4.8 km, in which very small particles are present, with maximum concentration of 120000 1/cm3 at the radius of 0.05 Pm. This layer seems to contain mainly the smut. In other layers, the maximum concentration is observed for particles with a radius size of 0.16 Pm. Depolarization of one-shot scattered radiation is stipulated by deviation of the particles’ shape from a spherical form. The D ratio in the depolarization layer of 2.5-3.5 km is practically constant and does not exceed 0.20-0.25 at a sharp increase of the backscattering ratio R at 3.0 km. It shows that the aerosol from Iraq is represented mainly by non-spherical particles. The vertical level-by-level cut shows a flaky structure of aerosol particles. At higher altitudes, from 3.5 km up to 4.8 km, a simultaneous increase of both D and R is observed. With larger particles in the layer of 4.8-5.4 km, i.e. with the increase of the

Influence of Atmospheric Aerosol Contamination on the Regional Climate

409

penetration depth of the exploring impulse into the aerosol, the depolarization ratio correlates with the backscattering ratio. Herewith the D values do not exceed 0.3. The concentration of small particles is 600 l/cm3, with a maximum for particles of 0.16 Pm radius. On the whole, the aerosol particles in all thick mass of the dust are not spherical, and their concentration has the tendency to decrease with height. The extinction coefficient at 532 nm reaches the maximum in a layer of 4.8-5.4 km (0.09 km-1), at an effective particle radius of 0.6 Pm. Nuclear explosions in the Lop-Nor range (May 15, 19, 1992) and underground nuclear explosions in India (May 1998) One of the causes of aerosol increase in the atmosphere is the radioactive burst in the epicenter of underground nuclear explosions. If after an explosion, the aerosol cloud gets into the zone of jet flow, it will settle down or rise up depending on the leg it encountered (anticyclone or cyclone). The higher the wind velocity in the jet is, the higher the layer of maximum values of aerosol concentrations is located. The upward aerosol transfer on a cyclone leg is more intensive than the downward transfer on an anti-cyclone leg, as the ascending vertical velocities are more powerful than the descending ones. After the underground nuclear explosion at Lop-Nor on May 15, 1992, an aerosol cloud 2 km thick (from 4.5 to 6.5 km) was observed, with the maximum aerosol concentration at 5.8 km (R=62.8) (Chen and Lelevkin, 2000). The fast exponential increase of the distance ratio up to the maximum value, followed by a linear decrease up to a height of 7.5 km was observed. By May 23 the thickness of the layer did not change, and the maximum of aerosol concentration moved to the height of 6.5 km (R=6.5). The maximum concentration decreased by 7 R units per diem, and moved upwards by 200 m per diem. After the underground nuclear explosion in India, in May, 1998, the aerosol particles were concentrated on and above the tropopause, before the aerosol layer of increased concentration came. On May, 14 the aerosol cloud 4.8 km thick was fixed with a maximum concentration at 12.4 km (R=23.6). The layer enveloped the upper troposphere and the lower stratosphere; the tropopause being diffused without any precise boundary of division. By May 17, the maximum of the layer moved lower, to 5.63 km, and the maximum of the distance ratio had reached a large value – Rmɚɯ=70.8. The aerosol cloud was located on the descendent anti-cyclone leg of the jet stream, with an average settling velocity – Rmɚɯ of 2.25 km per diem, and a velocity of diminution – R of 15.7 units per diem. Long-distance transport of aerosol Long-distance transport of aerosol in the troposphere is one of the relevant factors influencing the environmental change, both on the global and on the regional scale. The ecosystem of Central Asia is especially sensitive to this factor, wherever the free dominating offset of a natural aerosol from desertificated territories is observed in different atmosphere layers. The long-distance transport of aerosol has a strong impact on the radiative state of the atmosphere, the cloud-formation processes, the melting regime of glaciers and, in general, on the climate change in the region. For the analysis of the problems connected to the long-distance transport of air masses in the atmosphere, their movement trajectories should be learned first of all.

B. B. Chen and V. M. Lelevkin

410

For further analysis of the long-distance transport of aerosol pollution to the Tien-Shan territory, the backtrajectories (BT) of air transfer were developed at a level of the leading current 500 gPa, taking into account the regional physico-geographical conditions. All trajectories were classified into groups, first by the type of air masses (warm, cold and transforming), and then by six types of sources of natural pollution. Classification of back trajectories by sources of pollution (Chen et al., 2004) The backtrajectories highlight the basic sources of the aerosol pollution of the air masses, which can be assigned to two categories: one of natural and the other of anthropogenic character. The anthtopogenic sources include all large industrial centres, conglomerates, and sites of mining and processing of ores, coal and other natural raw materials. However, the greatest contribution to the aerosol pollution of the troposphere over the Central Asian region is made by the natural sources  vast territories of deserts, semi deserts, and steppes lying on the pathways of AM. These sources, classified by type, are: Type I.

Northern Africa (Sahara, Libyan Desert);

Type II.

Middle East (desert of Arabia, Iranian upland, Afghanistan);

Type ɒ.

Volga river basin, Western Kazakhstan (steppe, semi deserts, desert, Caspian Sea lowland, Ustyurt plateau, Aral Sea basin);

Type IV.

North-east Kazakhstan (Kazakh low hills, deserts of Betpak-Dala, Moyinkum, Taukum);

Type V.

Central Asia (Karakum Desert, Kyzylkumy Desert);

Type VI.

Southern Asia and China (Takla-Makan Desert).

Figure 3 shows the characteristic backtrajectories of air mass transfer and centers of the possible sources of aerosol pollution. The histogram of recurrence of the backtrajectories classified by pollution sources (Figure 4) clearly demonstrates the contribution of different centers of aerosol generation to the total air pollution over the Tien-Shan, in relation to the type of air mass. The contribution of individual groups of trajectories to the air transport to the studied region in a definite period of time was estimated from the relative number of the trajectories of this group, and by dividing the air masses into types. The proposed approach gives a unique solution of a problem of BT classification into types with characteristic physical propeties of AM and sources of pollution (Table 2). Thus, the most frequently observed trajectories are types II, III, IV. The trajectories transporting the dry and warm tropical air mass are referred to as types I, II, V, VI, and those with rather wet and cold air mass from moderate or arctic latitudes as type IV. In most cases the air flows transverse a rather long path over desert and steppe territories, so the extent of its transformation and pollution is very great. BT of types I and VI are seldom observed

Influence of Atmospheric Aerosol Contamination on the Regional Climate

411

Figure 3.

Diagram of the transfer of air masses for the different types of pollution sources.

Figure 4.

Histogram of recurrence of backtrajectories classified by sources of pollution and type of air mass.

Table 2.

Classification of backtrajectories by type of air mass and source of pollution. Classification of BT by source of by air mass pollution warm I cold warm II cold warm III cold warm IV cold warm V cold VI warm transformed cold ĺ warm

Recurrence, % out of type out of total cases 83.33 7.27 16.67 77.08 29.09 22.92 44.19 26.06 55.81 6.25 19.39 93.75 84.21 11.52 15.79 3.03 100.00 3.64 100.00

412

B. B. Chen and V. M. Lelevkin

Most of the long-distance transport of air to the studied region follows the trajectories of type I and the transformed ones, when in four days distances of several thousand kilometers are covered, starting from North Sahara or from the Atlantic and Western Europe. However, such trajectories are observed only in 7.27 and 3.64% of cases, respectively. Most frequently the air comes from the Middle East regions (type II) and the Caspian Sea lowland (type III). In total, the number of such trajectories amounts to about 55% of all cases. Almost each trajectory type can be divided in a subtype, according to the territory passed and the speed of air mass movement. The trajectories passing over several sources of aerosol pollution appeared most difficult to classify. For example, the trajectories of types I and II can pass over the territory of type V and, sometimes, type III. The trajectories of type III can pass over the regions of type IV and V. Thus, it is possible to determine a single center of pollution for trajectories IV, V and VI, but not for the trajectories I, II, III and transforming. For the latter group, the aerosol accumulation takes place over considerably large spaces, so the seasonal circulation and particular synoptic conditions describing the vertical component and the speed of air flow movement should be taken into account. Distribution of the aerosol optical and microphysical properties in the Atmospheric Brown Clouds A further study of the transport of aerosol pollution in the troposphere in relation to the actual optical-microphysical properties of aerosols, determined by the method of multiwave lidar sensing of atmosphere, allows the identification with high confidence of the various sources of aerosols. The effectiveness of the pollutants transport under equal conditions is determined by the lifetimes of pollutants in the atmosphere, which in turn depend on the character of the surface over which the air masses flow. The optical and microphysical properties of aerosol allowed the identification of the days during which the aerosol concentration in separate layers exceeded the background 102…103 times. Two such layers were identified: the Atmospheric Brown Cloud (Ⱥȼɋ) in the upper troposphere (5.59.1 km) and Ⱥȼɋ in the lower troposphere (1.74.3 km). The highest aerosol concentrations in the separate layers of the troposphere were observed most frequently in warm periods (May  September), and much less often in cold seasons. The optical properties in the Ⱥȼɋ change in the broad limits. The processes of transformation of the optical state of the troposphere are connected to the variation of the scattering properties of the aerosol in clouds, and hence to the microphysical properties of the aerosol. The atmospheric aerosol is a particle system of changing microphysical properties, composed of several fractions of different origin. Let's review the processes of transformation of a particle size distribution function of aerosol following a change of air masses, and compare the obtained quantitative assessment inside a single-type AMs. The important point is the selection of data applicable to such synoptic situations in which the aerosol generation can be connected to an air mass of one type only. The aerosol systems, similar to Ⱥȼɋ in the lower troposphere, will be generated by powerful convective processes in an AM of the continental moderate air. The fact that they are observed over different regions proves, as was pointed out above, that the sufficiently widespread physical phenomena and a process of the long-distance transfer take place over an

Influence of Atmospheric Aerosol Contamination on the Regional Climate

413

inversion layer. The appearance of clouds in our region, as a rule, is connected to AM from moderate latitudes (BT III) and is measured as localized areas of increased concentration of aerosol (R532!20, D532!0.3 km-1). The average values of the particle concentration per unit volume of air are 500700 Pm2/cm3. The vertical scale of similar formations ranges from 0.3 up to 1.5 km, with the lower borders at the altitude of h=1.73.5 km. The distributions of the aerosol surface area versus particle size in the clouds measured on 28.07.03 and 25.02.03, shown in Figure 5 (curves 1 and 2), indicate the dominating contribution of submicron (SBM) particles.

Figure 5.

Distributions of the aerosol surface area versus particle size, for ABC in various types of AM, in the lower (a) and upper (b and c) troposphere.

The concentration ratios between the modes show that the contribution of finedispersed and large fractions to the total balance does not exceed 15%. It corresponds to the known assumption that the dependencies of the dry precipitation speed on the particle size have minima in the submicron range. Thus, the elements bound to such particles are present in air longer than other elements, and can be transported for larger distances. The clouds observed in the lower troposphere when the inflow of arctic air (BT IV) dominates in the region are on average characterized by the same values of the optical parameters as the clouds in the previously reviewed case. However, the spectra of the distribution of the aerosol surface area versus particle size are noticeably shifted to the range of smaller particles (radiuses r=0.10.2 Pm), and are described by a broader logarithmically normal function (curves 3 and 4). The observation can be explained by the offset of aerosol containing a great amount of loes dust and products of anthropogenic pollution composed of outbursts from the industrial plants of Kazakhstan. In case of the pollution import by cool tropical air flows (BT II), the distribution of aerosol surface area versus particle size in the cloud generallly follows curves 3 and 4, but the dominating mode of the size spectrum occurs at the radius of r=0.15 Pm. The cloud composition is likely enriched with a saline component of a desert aerosol (r<0.4 Pm) and by loes dust. The probability of synoptic situations in which the clouds are formed and observed in the lower troposphere, is much higher in summer. Thus, one of the effects related to the longdistance transport of air is the transformation of the particle size ranges, related to the departure of different aerosol components from their sources. However, one should keep in mind that the processes aerosols take part in when transported to the atmosphere are complicated by the variability of such properties as the enrichment coefficient, the speed of dry precipitation and the distribution of aerosol mass versus particle size for different

414

B. B. Chen and V. M. Lelevkin

chemical elements. Analysis of the mechanisms of heterogeneous aerosols merging, of the power and spectrum of the sources and of the physico-chemical features and composition of the particles that determine the generalized gravity of the aerosol matter, can probably explain the results obtained from the measurements of the Ⱥȼɋ optical properties in the upper troposphere. The higher values of the scattering ratio at wavelength 1064 nm (Rmax!300) are characteristic for such clouds. They are explained by increased concentrations of coarsedispersed aerosol proved by the microphysical computing (the distributions of the aerosol surface area versus particle size shown in Figure 5), for a case of the pollution offset in the upper troposphere by air flowing along trajectories III (Figure 5b) and V (Figure 5c). The essential increase of the contribution of the fine-dispersed and submicron fractions in the total concentration of the particles forming a layer is notecible in the first case and that of the coarse-dispersed and SBM fractions in the second case. With this, the stability and poor responsiveness of the distribution of submicron particles versus their size to aerosol origin is well expressed, as well as the presence of large particles in quantities actually constant to different extents. According to all the evidence, the aerosol composition is more safely characterized not by the absolute values of the concentration, but by the concentration ratios for different fractions. The obtained data, analysed in terms of the distribution of absorbing elements versus particle size, show that the major part of the combustion products is accumulated on particles with radius r>0.5 Pm. It should be noted that there is not sufficient support for the assertion that the high layers contain the anthropogenic components in smaller amounts than the lower layers, related with more remote and primarily natural sources. The values of an imaginary part of a refraction index (k=0.020.04) are representative for Ⱥȼɋ both in the lower and in the upper troposphere. Thus, while studying the impact of the tropospheric aerosol pollution on the ecology of the atmosphere, soil, surface waters, glaciers and other major natural objects, the question about the place from where the air mass is transported to the region is doubtlessly relevant, as it determines the polluting rate of the air mass and the composition of the anthropogenic and natural impurities. Not less important is the question, where do these air masses travel to, as the answer determines the areas of further studies. References Bluth G.J.S., S.D. Doiron, C.C. Schnetzler, A.J. Krueger and L.S. Walter; Global tracing of the SO2 clouds from the June 1991 Mount Pinatubo eruption, Geophys. Res. Lett. 19 (1992) 151-154. Chen B.B. S.S. Khmelevtsov, V.A. Korshunov and A.M. Vdovenkov; Multiwavelength aerosol and Raman lidar. Proc. 21 JLRC, Quebec, Canada, 8-12 July 2002, 65-68. Chen B.B. and V.M. Lelevkin; Stratospheric aerosol layer over Central Asia. KRSU, Bishkek (2000). Chen B.B., L.G. Sverdlik and P.V.Kozlov; Optics and microphysics of atmospheric aerosol. KRSU, Bishkek (2004). Ivlev L.S., V.G. Sirota, S.N. Khvorostovsky; Volcanic sulfur dioxide oxidation influencing the concentration of sulfate aerosols and ozone in stratosphere, Optics of Atmosphere 3 (1990) 37-43. Okabe Kh.; Photochemistry of small molecules. Mir (1981). Stolypina N.V.; Seasonal changes of circulation intensity in stratosphere of northern hemisphere. Gidrometeoizdat (1981). Toktomyshev S. Zh. and V.K. Semenov; Ozone holes and the climate of mountain regions in Central Asia. Istanbul (2001). Turko R.P., O.B. Toon and R.C. Written; The 1980 eruption of Mount St. Hellens: physical and chemical processes in the stratospheric clouds. J. Geophys. Res. 88 (1983) 5299-5319.

Assessment of Air Pollution and Ecosystem Buffer Capacity in the Industrial Regions of Ukraine Mykola M. Kharytonov1 Larisa B. Anisimova, Natalia P. Gritsan, and Andriy P. Babiy 2 1

State Agrarian University, Voroshilov st.25, Dnepropetrovsk, 49600, Ukraine 2 Institute for Nature Management Problems and Ecology, N.A.S.U. Key Words: Air pollution, Heavy metals, Ecosystems, Buffer capacity

Abstract Investigations were conducted at urban and rural sites of the Dniepropetrovsk Province (Ukraine) during the last decade, and were aimed at an assessment of the air pollution and the buffer capacity of local ecosystems. The following themes were considered in the study: the regional distribution of industrial emissions, the structure of the ecosystems, an estimation of the buffer capacity of the soil and vegetation cover. Introduction The air basin of large industrial cities of the south-eastern part of Ukraine is extremely polluted in spite of the permanent convection and transfer across boundaries. High levels of atmospheric pollution is the result of various economic activities, of which industry and transportation are the most polluting ones. The total emission volume from enterprises of the Donetsk - Prydniprovya industrial regions remains rather high and makes up 8 –15 % of the total volume of emissions in the country. Especially high volumes of emissions were recorded in the cites of Kryvyi Rig (10.1%), Mariupol (7.8%), Donetsk (5.0%), Zaporizhzhia (3.3%), Dneprodzerzhynsk (2.9%), Enakiyeve (2.6%), Dnepropetrovsk and Alchevsk (2.4%), Lugansk (2.3%), and Debaltseve (2.2%) (MEPNSU, 1998). The high concentration of the industries in large cities aggravates the very negative impact of emissions on natural ecosystems and human health (Report, 1997; MEPNSU, 1998). The Dnepropetrovsk Region holds one of the first places in Europe with respect to air pollution. Major stationary sources of emissions are: centralized heat and power production (33%), metallurgical industries (25%), coal mining (23%), and chemical and petroleum industries 2% (Kharytonov et al., 2002). This paper presents a part of complex investigations concerning an assessment of the environmental impacts of air pollution in the south-eastern part of Ukraine. Materials and methods The investigations were conducted at urban and rural sites of the Dnepropetrovsk Province during the last decade of the 20th century. Dnepropetrovsk Province is situated in the southern part of Ukraine, on both banks of the river Dnepr. It occupies an area of 31923 sq km. (5.3% of the total territory of Ukraine), which is comparable with the territory of the Netherlands (Babiy et al., 2003). The Province has 22 administrative districts (rural areas) and 13 municipalities. The capital and the largest city is Dnepropetrovsk, with an area of 378.5 sq km inhabited by 1.2 million people. 415 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 415–420. © 2006 Springer. Printed in the Netherlands.

416

M. M. Kharytonov et al.

Dnepropetrovsk Province has a wealth of natural resources including iron, manganese, scandium and uranium ores, black coal, water resources, very fertile soils and favourable climatic conditions for the ecosystems development. Natural recourses formed the basis for the leading industries at the national and regional level. Two of the metal-ore reserves are among the largest in the world: the iron ore reserve in Krivoy Rog (Krivbass) and the manganese ore reserve in the vicinity of the city of Nikopol. The conditions of soil formation in the Province are consistent with the geographical zonality. The main type of soil is black soil – chernozem. About 80 % of the territory is occupied by chernozems of different subtypes. The black soil fertility is famous in the world. An atomic absorption spectrophotometer has been applied for heavy metal determination in soils. Concentrations of potassium and sodium were determined with a flame photometer. For the determination of soil fluorides, an ion-selective electrode was used. Statistical methods were applied to describe quantitatively the relationships between industrial emissions and other components of the ecosystems. Results and discussion The average annual emission load from industries and transport ranged from 109 to 25 tons per square kilometre, in 1980 and at the end of the century, respectively (Babiy et al., 2003). Gas, liquid and solid wastes are major sources of emissions, and causes of the environmental degradation in the Dnepropetrovsk province. The main factors which determine the air quality are the volume and composition of industrial emissions. Table 1 shows the average annual emissions of airborne pollutants, calculated for the districts of the Dnepropetrovsk province. Taking these data into account, it is clear that the soil covering can be extremely damaged with airborne pollutants from the facilities of metallurgical, power, mining, chemical and petrochemical industries. The level and character of air pollution depend on many factors: type, quantity and quality of industrial emissions, type and intensity of physical and chemical atmospheric processes, land surface etc. The total area of wasted land in the Province is about 340000 sq km; the highest in Ukraine. The mining industry is the main source of land-devastating pollution. About 100 years of intensive exploitation of the iron ore basin in Krivoy Rog led to the accumulation of about 2 billion cubic meters of wastes from enrichment processes. Airborne emissions in the Province are produced by the network of iron, manganese and scandium ore enriching enterprises. In addition, there are 8 tailing ponds in Krivbass, which occupy an area of 90 sq km. The coal mining wastes, which occupy 0.2 sq km of land, amount to about 50 million tons. The metallurgical industry is another large source of waste in the Province – more than 11 million tons of blast-furnace slack occupies 2.3 sq km of land. More than 8 millions tons of ferro-alloy slack occupies an area of 0.1 sq km. About 10 million tons of mixed chemical wastes have been stored in an area of 6.2 sq km, and 60 million tons of ash from the Dnepropetrovsk power station in an area of about 8.5 sq km. Humans have been intensively changing the land. The landscapes of the Ukrainian steppe zone has changed almost completely. Only 0.8 % of the Dnepropetrovsk Province remains as a more or less natural ecosystem. The prevalence of highly fertile soils and favourable climatic conditions favoured a very high average level of agricultural development of the land; 79 % on average (more than 25000 sq km), and up to 90 % locally. The share of arable lands varies from 55 to 80%. The increasing area of agricultural land (especially the

Assessment of Air Pollution and Ecosystem Buffer Capacity

417

arable land) led to the diminution of natural steppes, forests and other vegetation. It should be kept in mind that natural areas determine the biosphere stability, agricultural production, water exchange, cycles of substances and energy flows. Now, the forests occupy only about 4% of the Province, and this number should be increased up to an optimal level of 8 %, while the percentage of arable land should be decreased. Average annual emissions in districts of the Dnepropetrovsk province. Emission of individual substances thousand tons

District

Total emission thousand tons

Gas and liquid

Solid

SO2

CO

Nx Ox

CH4

VOC

Other gases and liquids

Sub-total emissions thosand tons

Number of enterprises

Table 1.

Apostolovsky Vasilkovsky Verkhnedneprovsky Dnepropetrovsky Krivorozhsky Krinichansky Magdalinovsky Mezhevskoy Nikopolsky Novomoskovsky Pavlogradsky Petropavlovsky Petrikovsky Pokrovsky Pyatikhatsky Sinelnikovsky Solonyansky Sofievsky Tomakovsky Tsarichansky Shirokovsky Yurievsky

16 8 15 210 99 6 7 5 58 27 46 4 16 22 36 19 8 8 16 7 4 2

141.4 0.08 0.49 249.2 368.3 0.05 2.65 0.155 47.60 0.88 6.795 0.131 6.139 0.685 5.14 0.672 0.163 0.075 0.199 0.132 0.028 0.011

98.46 0.06 0.42 192.8 300.3 0.04 2.63 0.21 39.65 0.64 5.42 0.06 4.31 0.28 3.69 0.537 0.112 0.062 0.071 0.106 0.014 0.005

42.93 0.03 0.06 56.35 67.74 0.01 0.015 0.032 5.956 0.242 1.647 0.072 1.831 0.409 1.45 0.135 0.051 0.013 0.128 0.026 0.014 0.006

65.15 0.006 0.15 72.82 27.13 0.002 0.003 0.066 0.812 0.178 2.746 0.033 2.977 0.021 2.58 0.079 0.043 0.025 0.004 0.002 0.003 0.001

7.60 0.019 0.08 84.78 256.7 0.01 0.02 0.046 36.42 0.125 1.595 0.016 0.832 0.14 0.434 0.269 0.014 0.029 0.028 0.075 0.005 0.003

25.68 0.011 0.12 29.73 12.94 0.007 0.005 0.1 2.02 0.30 0.94 0.009 0.44 0.08 0.461 0.087 0.026 0.007 0.015 0.021 0.006 0.001

0.001 0.017 0.164 0.316 0.005 2.556 0.001 0.028 0.001 0.027 0.008 0.003 0.053 0.03 0.001 0.001 0.001 -

22.91 0.039 38.48 3365.57 1175.59 1.94 0.466 142.66 32.12 75.37 10.10 11.98 52.40 30.78 26.54 0.30 16.68 2.73 0.001 0.052

0.004 0.019 0.014 1.075 2.30 0.01 0.22 0.002 0.038 0.001 0.037 0.018 0.104 0.068 0.002 0.006 0.004 -

Total for province

639

830.9

649.9

179.2

174.8

389.2

72.98

3.186

50067.7

4.83

On one hand, humus is the first factor of soil fertility, but on the other hand, it can accumulate various hazardous substances, which can reduce the soil quality. A strong correlation between the contents of humus and most heavy metals in the soils has been found (Table 2).

M. M. Kharytonov et al.

418 Table 2.

Correlation between content of humus and heavy metals. Pb

Cd

Ni

Cr

Fe

Mn

Cu

Zn

0.80

0.67

0.87

0.83

0.46

0.66

0.76

0.02

Oxides and hydroxides of Fe and Mn are ordinary components of black soils, but their impact on the behaviour of microelements is very important. These compounds can absorb microelements, because they form membranes in soil (Kabata-Pendias et al., 2003). High content of Fe and Mn oxides and hydroxides in soil may lead to significant changes in the geochemical balance. However, the environmental impacts are not equally distributed over the territory. It is a well-known fact, that impacts in certain industrial areas are higher than in others. Table 3 shows the concentrations of 15 elements in urban soils. The maximal concentration of Fe (on average 10 times higher than normal) was observed in the soils of the Krivoy Rog city, because of the high natural background level of Fe in the area, which is a natural reserve of iron ore, as well as huge industrial emissions. Phosphatase and urease activity was shown to have a strong negative correlation with the distance from the mine and with the concentrations of Cu and Pb in the dust from iron ore mining, while invertase activity had a rather weak correlation (Ruth and Kharytonov, 2003). Table 3. Element Ca Mg K Na Al Fe Mn Zn Cu Co Ni Pb Cd Cr F

Content and coefficient of variation (CV) of contaminants in soils of the Dnepropetrovsk Region (mg/kg dry mass). Dnepropetrovsk Content CV 23644 58,8 3589 37.7 10412 26.8 4100 41.7 27372 29.6 27913 107.5 791 83.2 126 45.8 29 61.8 11 20.4 16 27.7 27 57.7 1 22.1 42 29.6 387 31.5

Krivoy Rog Content CV 95250 150.7 8150 44.0 9875 40.0 2950 64.0 31200 52.9 62675 84.8 733 34.8 85 68.2 17 20.7 14 25.2 48 47.4 21 25.0 6 142.7 48 13.9 554 58.6

Dneprodzerzhinsk Content CV 128800 87.9 4600 37.6 6470 58.4 4240 33.3 25300 43.7 55033 97.3 1563 47.7 177 75.6 44 67.3 13 34.5 22 46.3 45 89.5 2 93.2 48 25.9 715 43.9

Rural areas Content CV 15421 37.5 4884 42.6 13189 18.4 4242 23.8 33613 34.2 20643 25.2 662 26.3 70 41.9 18 24.4 12 12.2 19 26.8 16 35.3 1 14.4 41 18.0 320 32.5

Especially high contents of Cu were determined in the soils of the cities of Krivoy Rog and Dnepropetrovsk. Copper is a very immobile soil element. However, the amount of Cu in the atmosphere, soils and plants appeared to increase during last years. Copper is one of the most polluting components in the Dniepropetrovsk Region. The content of zinc in the soils of Dnepropetrovsk and Krivoy Rog was on average 3 times higher than its background level, so it seems to pose another environmental problem.

Assessment of Air Pollution and Ecosystem Buffer Capacity

419

Concentrations of chromium in the soils of large cities were on average 4 times higher than normal. More attention should be paid to this element. The nickel content in big industrial centres and surrounding rural areas was 5-7 times and 3-5 times higher than the background level, respectively. Lead is an element ranked first class in danger, according to the Ukrainian classification. Among the heavy metals, Pb is not a very mobile element. The concentrations of Pb determined in soils were generally higher than normal and higher in the cities than in rural areas. Fluorides also contaminated the soils of the studied territories. The highest level of soil contamination with heavy metals coincides with other information on industrial emissions in the cities of Krivoy Rog, Dnepropetrovsk, and Dneprodzerzhinsk, because of the high concentration of environmentally dangerous industries located directly in these cities. Thus, the level and character of soil pollution depended mainly on the quantity and quality of industrial emissions and the type of industries. It was found that iron, copper, zinc, lead and fluoride were most important among the contaminants in terms of airborne soil contamination. Corresponding air and soil samples were taken at several sites in Dniepropetrovsk city and analysed for the heavy metal content. Table 4 shows the calculated correlation between the content of heavy metals in the air and in the soil samples. Correlation between contents of heavy metals in corresponding air and soil samples from Dniepropetrovsk city.

Table 4.

Manganese

Iron

Nickel

Copper

Zinc

Lead

0.65

0.99

0.95

0.98

0.99

0.95

Calculations were also made to predict the ecosystem buffer capacity for fluoride. The distribution of forests and agricultural land within the Dnepropetrovsk province was taken into account to assess the ecosystems biological potential for F reutilization (Table 5). Concluding remarks High levels of air, soil and plant pollution is the result of various economic activities, of which industry is a very important contributor. The spectra of pollutants which affected urban soils depended mainly upon the kind of industry. The negative impact of industry has led to changes in the chemical composition of the biosphere, and the accumulation of pollutants in soils, plants, etc. Consequently, the food chain has been contaminated as well, which has had a negative impact on human health. References Babiy, A.P., M. Kharytonov and N.P. Gritsan; Connection between emissions and concentrations of atmospheric pollutants, in D. Melas, D. Syrakov (eds), Air Pollution Processes in Regional Scale. NATO Science Series, IV: Earth and environmental sciences, Kluwer Academic Publishers, Dordrecht (2003) 11-19. Kabata-Pendias, A. and H. Pendias; Trace elements in soils and plants, 2nd ed., Levis Publ., Boca Raton (1987). Kharytonov, M., N. Gritsan and L. Anisimova; Environmental problems connected with air pollution in the industrial regions of Ukraine, in: I. Barnes (ed.), Global atmospheric change and its impact on regional air quality. NATO Science Series, IV: Earth and environmental sciences, Kluwer Academic Publishers, Dordrecht (2002) 215-222. Kharytonov, M., A. Zberovsky, A. Drizhenko and A. Babiy; Blasting impact assessment of the dust-gas clouds

420

M. M. Kharytonov et al. spreading in the iron ore mining region of Ukraine, Data fusion for situation monitoring, incident alert and response management. NATO ASI, (2003) (CD-ROM).

MEPNSU; National report on the environmental conditions in Ukraine for 1998. Ministry for Environment Protection and Nuclear Safety of Ukraine (1998), (CD-ROM publication, www.grida.no). Report; National report on environment state of Ukraine in 1995. Raevsky Press, Kiev (1997) (in Ukrainian). Ruth, E. and M. Kharytonov; Integrated approach for Assessment of Polluted Areas, in: R.O. Rasmussen, N.F. Koroleva (eds), Social and environmental impacts in the North: Methods in evaluation of socioeconomic and environmental consequences of mining and energy production in the Arctic and SubArctic. NATO Science Series, IV: Earth and environmental sciences. Kluwer Academic Publishers, Dordrecht (2003) 57-66.

Table 5.

Ability of forested and arable areas in the Dnepropetrovsk province to accumulate fluorine.

Districts Apostolovsky Vasilkovsky Verkhnedneprovsky Dnepropetrovsky Krivorozhsky Krinichansky Magdalinovsky Mezhevskoy Nikopolsky Novomoskovsky Pavlogradsky Petropavlovsky Petrikovsky Pokrovsky Pyatikhatsky Sinelnikovsky Solonyansky Sofievsky Tomakovsky Tsarichansky Shirokovsky Yurievsky Total for province

Land area

Land ability to accumulate fluorine

thousand ha

tons

Forested

Arable

Forested

Arable

2.2 2.2 12.0 5.9 3.8 1.8 2.3 3.1 3.1 9.3 8.9 2.7 2.2 3.4 6.9 2.7 3.3 1.8 1.9 6.4 1.5 2.3 89.7

104.428 119.002 86.330 110.864 102.185 149.414 137.842 112.104 133.492 147.182 113.411 108.697 56.017 106.358 141.792 144.472 152.905 121.224 94.839 73.409 96.729 79.502 2492.198

44 44 240 118 76 36 46 62 62 186 178 54 44 68 138 54 66 36 38 128 30 46 1794

208.9 238.0 172.7 221.7 204.4 298.8 275.7 224.2 267.0 294.4 226.8 217.4 112.0 212.7 283.6 288.9 305.8 242.4 189.7 146.8 193.5 159.0 4984.4

Two Neural Network Methods in Estimation of Air Pollution Time Series Hikmet Kerem Cigizoglu1, Kadir Alp2, and Müge Kömürcü2 Division of Hydraulics, e-mail:[email protected], 2 Environmental Engineering Department, Civil Engineering Faculty, Istanbul Technical University, Maslak, 34469, Istanbul, Turkey

1

Key Words: Particulate matter, Neural networks, Feed forward back propagation, Radial basis functions, Estimation

Abstract The measurement of air pollution parameters is a costly process. Due to several reasons, the devices may not take measurements for certain days. In such cases robust estimation methods are quite necessary in order to fill the gaps in the time series. Artificial neural networks have been employed successfully for this purpose for hydrometeorological time series, as reported in literature. In this study, modelling of the time series of air pollution parameters was investigated using two ANN methods; a radial basis function algorithm (RBF) and feed forward back propagation method (FFBP). The ANN methods were employed to estimate the PM10 values using the NO and CO values. The data were from a measurement station in Istanbul, Turkey. The results of an initial statistical analysis were considered in the determination of the input layer node number. In the estimation study, values corresponding to other air pollution parameters were included in the input layer. The results were compared to those obtained with a conventional multi-linear regression (MLR) method. Introduction Air pollution is an extremely significant issue that should be focused on all around the globe, as it affects human health, ecosystems and the environment, harms materials, buildings, and in turn the economy. It is defined as a condition, in which substances that originate from natural and anthropogenic activities are present in the air at concentrations sufficiently high above their normal ambient levels to produce considerable impacts on humans, animals, vegetation, or materials (Seinfeld and Pandis, 1998). Not all the studies consider air pollution to be a consequence of both natural and anthropogenic activities, some studies define air pollution only in terms of anthropogenic origin. Various problems of human health, including respiratory problems, hospitalization for heart or lung diseases, and premature deaths are caused by air pollution. Children, who spend more of their time outdoors while their lungs are still developing, are more susceptible to the risk. Elderly people and people with heart or lung diseases are also sensitive to several types of air pollution (U.S. EPA, 2002). Not only regional sources generate pollution, but also transport processes induced by weather patterns. It is important to detect the dominant processes, to take the proper action. If transport processes are dominating, imposing regional standards on industrial sources and limiting the anthropogenic activities will not be adequate to prevent the pollution. Regional sources of air pollution comprise several sources categorized as natural and anthropogenic. The natural sources are windblown dust, volcanoes, ocean sprays, evaporation and wildfires. 421 I. Barnes and K. J. Rudzinski (eds.), Environmental Simulation Chambers: Application to Atmospheric Chemical Processes, 421–431. © 2006 Springer. Printed in the Netherlands.

422

H. K. Cigizoglu et al.

One of the most important sources of natural pollutants is volcanic eruptions. During a volcanic eruption, huge amounts of gases and particulate matter are released into the atmosphere. Sulphur oxides (SOx) belong to the major constituents of these gases. The ultraviolet radiation coming to the Earth is scattered and absorbed by volcanic particles and gases (Wayne, 2000). Another natural source of pollution is ocean sprays. Sprays of salt from the oceans and seas are transferred to the atmosphere by evaporation and winds. Evaporation is another significant source of natural air pollution. Numerous major gases are transferred to the atmosphere via evaporation from water bodies with large surface areas, such as oceans. In the development phase of the plankton, oceans become the sources of CO2 emission. Furthermore, forests release hydrocarbons by photochemical reactions. Although biomass burning can occur naturally, recently most of the biomass burning is done intentionally. Some of the intentional purposes are to control pests, to produce energy, to clear forested areas, and to mobilize nutrients. Major species emitted during biomass burning are CO, CH4, CH3Br, CH3Cl, NOx, N2O, sulphur compounds, volatile organic carbon species including non-metal hydrocarbons, PAH and particulate matter. In dry seasons, biomass burning is intense enough to be seen from outer space. Biomass burning also increases the ozone levels in the troposphere. Besides the natural sources, there are also anthropogenic sources of pollution, which are divided into two categories: stationary and mobile. Some stationary sources are factories, power plants, smelters, dry cleaners and degreasing operations; whereas some mobile sources are cars, buses, planes, trucks, and trains (U.S. EPA, 2002). Air pollutants are also categorized, like the pollution sources. The two categories are primary and secondary pollutants. Primary pollutants are the chemicals that are directly emitted to the atmosphere from known sources, while secondary pollutants are the species that are formed by the chemical reactions of primary pollutants (Wayne, 2000). Primary pollutants include sulphur oxides (SOx), nitrogen oxides (NOx), carbon monoxide (CO), hydrocarbons (HC), and particulate matter (PM). One of the most important secondary pollutants is ozone. The artificial neural network (ANN) approach seems to be a useful alternative for modelling the complex hydro-meteorological time series. The artificial neural networks, which are non-linear black box models, became standard tools in forecasting the rainfall (Bodri and Cermak, 2000; Luk et al., 2000; Sahai et al., 2000; Silverman and Dracup, 2000), in forecasting the stream flows and suspended sediments (Cigizoglu, 2003a; Cigizoglu, 2003b; Cigizoglu, 2004; Cigizoglu and Kisi, 2005), in air pollution research (McKendry, 2002; Cigizoglu et al., 2004), in ozone concentration studies (Narasimhan et al., 2000; Gardner and Dorling, 2001; Abdul-Wahab and Al-Alawi, 2002), in retrieval of temperature profiles (Kuligowski and Barros, 2001; Aires et al., 2002), and in many other scientific fields. In the majority of these studies the feed forward error back propagation method (FFBP) was employed to train the neural networks. The performance of FFBP was found superior to conventional statistical and stochastic methods in continuous flow series forecasting (Cigizoglu 2003a; Cigizoglu 2003b). However the FFBP algorithm has some drawbacks. It is very sensitive to the selected initial weight values, so its performance may vary significantly. Another problem within the application of FFBP is the local minima issue. Maier and Dandy (2000) summarized the methods used in the literature to overcome the local

Two Neural Network Methods in Estimation of Air Pollution Time Series

423

minima problem, such as training a number of networks with different initial weights, the online training mode to help the network to escape local minima, addition of random noise, employment of second order (Newton algorithm, Levenberg-Marquardt algorithm) or global methods (stochastic gradient algorithms, simulated annealing). Other ANN methods such as conjugate gradient algorithms, radial basis function, cascade correlation algorithm and recurrent neural networks were described by the ASCE Task Committee (2000). In the present study, two ANN methods – the FFBP with the Levenberg-Marquardt algorithm and the radial basis functions (RBF) – were employed to estimate the air pollution parameters measured at a station in Istanbul on chosen episode days, with the focus on the particulate matter. The results were compared with those obtained with the multi-linear regression (MLR) method. Description of ANN methods The feed forward back propagation (FFBP) The FFBP distinguishes itself by the presence of one or more hidden layers, whose computation nodes are correspondingly called hidden neurons of hidden units. The function of hidden neurons is to intervene between the external input and the network output in some useful manner. By adding one or more hidden layers, the network is enabled to extract higher order statistics. In a rather loose sense, the network acquires a global perspective despite its local connectivity due to the extra set of synaptic connections and the extra dimension of NN interconnections (Hagan and Menhaj, 1994). Figure 1 depicts the structure of a FFBP neural network. The ability of hidden neurons to extract higher order statistics is particularly valuable when the size of the input layer is large. The source nodes in the input layer of the network supply respective elements of the activation pattern (input vector), which constitute the input signals applied to the neurons (computation nodes) in the second layer (i.e. the first hidden layer). The output signals of the second layer are used as inputs to the third layer, and so on for the rest of the network. Typically, the neurons in each layer of the network have as their inputs the output signals of the preceding layer only. The set of the output signals of the neurons in the output layer of the network constitutes the overall response of the network to the activation patterns applied by the source nodes in the input (first) layer. The FFBP was trained using the Levenberg–Marquardt optimization technique. This optimization technique is more powerful than the conventional gradient descent techniques (Hagan and Menhaj, 1994). The FFBP algorithm has two significant drawbacks. Firstly, the performance of training simulations depends on different random weight assignments in the beginning of each training simulation. The FFBP network can be trapped by different local error minima each time, and the desired error value may not be attained. Secondly, FFBP can generate negative numbers when estimating small parameters. In the present study, different numbers of hidden layer nodes were tried, and the configuration that gave the minimum mean square error (MSE) was determined.

424

H. K. Cigizoglu et al. Input layer

Hidden layer

Output layer O1

X1 X2

O2

X3

O3

Xk

Figure 1.

Om

Structure of a feed forward neural network (FFBP).

The radial basis function-based neural networks (RBF) Figure 2 depicts the structure of RBF neural networks. RBF networks were introduced into the neural network literature by Broomhead and Lowe (1988). The RBF network model imitates the locally tuned response observed in biological neurons. Neurons with a locally tuned response characteristic can be found in several parts of the nervous system, for example, in cells in the visual cortex sensitive to bars oriented in a certain direction or other visual features within a small region of the visual field. These locally tuned neurons show the response characteristics bounded to a small range of the input space. The theoretical basis of the RBF approach lies in the field of interpolation of multivariate functions. The objective of





N

interpolating a set of tuples x s , y s s 1 with xs  Rd is to find a function F: Rd Æ R with F(xs) = ys for all s = 1, ..., N where F is a function of a linear space. In the RBF approach the interpolating function F is a linear combination of basis functions N

F ( x)

¦ w I  x  x  px s

s

(1)

s 1

where

˜

denotes the Euclidean norm, w1, ..., wN are real numbers, I is a real valued d

function, and p  –n is a polynomial of degree at most n (fixed in advance) in d variables. The interpolation problem is to determine the real coefficients w1, ..., wN and the polynomial d D term p ¦l 1 a1 p j , where pl, ..., pD is the standard basis of –n and a1, ..., aD are real coefficients. The interpolation conditions are



F xs

ys,

s 1, ..., N

(2)

Two Neural Network Methods in Estimation of Air Pollution Time Series

425

and N

¦ w p x s

s

j

0,

j

1, ..., D .

(3)

s 1

The function I is called a radial basis function if the interpolation problem has a unique solution for any choice of data points. In some cases the polynomial term in Eq. (1) can be omitted and by combining it with Eq. (2), we obtain

Iw

y

(4)

where w = (w1, …, wN), y = (y1, …, yN), and I is an N x N matrix defined by

I

I  x

k

 xs



(5)

k , s 1, ..., N

Provided the inverse of I exists, the solution w of the interpolation problem can be explicitly calculated, and has the form: w = I-1 y. The most popular and widely used radial basis function is the Gaussian basis function

I x  c



e



 x  c / 2V 2



(6)

with a peak at center c  Rd, and decreasing as the distance from the centre increases.

Input layer

Figure 2.

Hidden layer

Output layer

Structure of a Radial Basis Function Neural Network (RBF).

The solution of the exact interpolating RBF mapping passes through every data point (xs, ys). In the presence of noise, the exact solution of the interpolation problem is typically a function oscillating between the given data points. An additional problem with the exact interpolation procedure is that the number of basis functions is equal to the number of data points, so calculating the inverse of the N x N matrix I becomes intractable in practice. The interpretation of the RBF method as an artificial neural network consists of three layers: a layer of input neurons feeding the feature vectors into the network; a hidden layer of RBF

426

H. K. Cigizoglu et al.

neurons, calculating the outcome of the basis functions; and a layer of output neurons, calculating a linear combination of the basis functions. The RBF method provides a unique solution in a simulation with constant spread values, unlike the FFBP. Similarly to FFBP, the RBF can provide negative estimations for measurements with low values. Both the FFBP and the RBF algorithms have short training time. However, multiple FFBP simulations are required to obtain satisfactory performance criteria, so the overall training is longer than with the unique RBF application. Different numbers of hidden layer neurons and spread constants were tried in this study. The number of hidden layer neurons that gave the minimum mean square error (MSE) was determined. The spread that gave the minimum MSE was also found simply by adding some trial-and-error loops in the program codes.

400

500

(a)

350 300

300 NO

PM10

400

200

(b)

250 200 150 100

100 0 0

50

100

150

200

50 0 0

50

Time (days)

100

150

200

Time (days)

8000

(c)

7000 6000

CO

5000 4000 3000 2000 1000 0 0

50

100

150

Tim e (days)

Figure 3.

The observed air pollution records (January 1999-April 2003).

Data In the study, the PM10, NO and CO measurements taken in the European part of Istanbul, Alibeyköy, between January 1999 and April 2003 were used (Figure 3). One measurement was taken per day. The data are not continuous and contain gaps. The common

Two Neural Network Methods in Estimation of Air Pollution Time Series

427

measurement set for the PM10 - NO pair covers 167 daily values, whereas there are 133 values for the PM10 - CO pair. Since the study was aimed at the estimation of PM10 values basing on other pollution parameters, the cross-correlations were computed in order to determine the time interval significant for the estimation process. Table 1 shows that only NO and CO values for times t and t-1 have considerable correlations with PM10 at time t (r1 and r2). First-, second- and third-order cross-correlations.

Table 1. Data pair

r1

r2

r3

Measurement period

PM10-NO

0.75

0.42

0.09

January 1999-April 2003 (167 days)

PM10-CO

0.57

0.30

-0.02

January 1999- April 2003 (133 days)

ANN application results PM10 estimation using NO In this part of the study, the FFBP and RBF methods were employed to estimate PM10 using the past and present NO measurements. The input layer consisted of two NO measurements at times t and t-1 since only the cross-correlations for two lags between PM10 and NO were significant (Table 1). The output layer had the unique PM10 value for time t. A hidden layer with 3 nodes was found adequate for FFBP, whereas the RBF provided best estimations for a spread equal to 61.9. The first 117 input-output patterns were considered for the training stage and the last 50 for the testing stage. The results of estimation are depicted in the form of time series and scatter plot in Figures 4 and 5 for RBF and FFBP, respectively. It was clear that FFBP performed better than RBF. The performance evaluation criteria (mean square error, MSE, and determination coefficient, R2) are given in Table 2. The multi-linear regression, MLR, was selected for comparison and its estimation results for the testing period are presented in Figure 6 and Table 2. It is obvious that both ANN methods provided performance significantly superior to MLR. Table 2.

ANN estimation results.

Case

Evaluation criteria

RBF

FFBP

MLR

PM10 estimation using NO

MSE

1408

1338

1617

R2

0.72

0.69

0.36

MSE

2232

2129

2409

R2

0.43

0.42

0.36

PM10 estimation using CO

H. K. Cigizoglu et al.

428 400

y = 0.989x + 17.122 R2 = 0.7163

350

Estimated PM10

PM10

- - - - RBF _____ Measured

400 350 300 250 200 150 100 50 0

300 250 200 150 100 50

0

10

20 30 Time (day)

40

0

50

0

20

40

60

80

100

120

140

Measured PM10

Estimation of PM10 using NO with RBF.

Figure 4.

350

350

- - - - FFBP _____ Measured

y = 0.8492x + 33.388 R2 = 0.6946

300

Estimated PM10

300

PM10

250 200 150 100

200 150 100 50

50 0

0

0

Figure 5.

10

20 30 Time (day)

0

40

20

40

60

80

100

120

140

100

120

140

Measured PM10

Estimation of PM10 using NO with FFBP.

250

- - - - MLR _____ Measured

350 300

y = 0.3022x + 68.081 R2 = 0.3634

Estimated PM10

200

250 PM10

250

200 150 100

150

100

50

50 0

0

0

Figure 6.

20 Time (day)

40

0

20

Estimation of PM10 using NO with MLR.

40

60

80

Measured PM10

Two Neural Network Methods in Estimation of Air Pollution Time Series

429

PM10 estimation using CO The study was repeated to estimate PM10 but this time using the past and present CO measurements. The input layer again consisted of two CO measurements (times t and t-1) and the unique output layer node represented the PM10 value to be estimated for time t. A hidden layer with 3 nodes for FFBP and an RBF network with a spread equal to 40.8 provided best estimations. However, the order of training and testing periods was reversed this time. Accordingly the first 83 input-output patterns were considered for training and the last 50 for testing stage. The estimation results for the testing period are depicted in the form of time series and scatter plot in Figures 7 and 8 for RBF and FFBP, respectively. It was seen that FFBP performed better compared with RBF in terms of MSE, but the R2 value for RBF was higher than FFBP (Table 2). Both ANN methods were significantly superior to MLR (Figure 9 and Table 2).

300

- - - - RBF _____ Measured

350

Estimated PM10

250 PM10

y = 0.354x + 64.662 R2 = 0.4252

250

300

200 150 100

200 150 100 50

50 0

0

0

0

20 Time (day)

20

40

40

60

80

100

120

140

Measured PM10

Estimation of PM10 using CO with RBF.

Figure 7.

250

350 - - - - FFBP _____ Measured

250 PM10

y = 0.3022x + 68.081 R2 = 0.3634

200

Estimated PM10

300

200 150 100

150

100

50

50 0

0 0

Figure 8.

10

20 30 Time (day)

40

50

0

Estimation of PM10 using CO with FFBP.

20

40

60

80

Measured PM10

100

120

140

H. K. Cigizoglu et al.

430 250

- - - - MLR _____ Measured

350 300

Estimated PM10

250 PM10

y = 0.3022x + 68.081 R2 = 0.3634

200

200 150 100

150

100

50

50 0 0

20 Time (day)

40

0 0

20

40

60

80

100

120

140

Measured PM10

Figure 9.

Estimation of PM10 using CO with MLR.

Conclusions In the presented study the air pollution parameter estimation based on two ANN methods, RBF and FFBP, was examined. Both methods provided significantly superior estimations compared with the conventional MLR method. The correlation statistics were employed successfully to determine the input layer node number. The performance evaluation criteria of FFBP for the testing period were slightly better than those of RBF. However, multiple FFBP simulations were required until obtaining satisfactory performance criteria, and this total duration was longer than the unique RBF application. The air pollution series may have gaps during the observation period. The presented study has shown that both ANN methods could be successfully employed to fill the gaps in the air pollution record. References Abdul-Wahab, S A. and S. M. Al-Alawi; Assessment and prediction of tropospheric ozone concentration levels using artificial neural networks, Environ. Model. Software 17 (2002) 219-228. Aires, F., A. Chedin, N.A. Scott and W.B. Rossow; A regularized neural net approach for retrieval of atmospheric and surface temperatures with the IASI Instrument. J. Appl. Meteorol. 41 (2002) 144-159. ASCE Task Committee; Artificial neural networks in hydrology I, J. Hydrol. Eng. 5 (2000) 115-123. Bodri, L. and V. Cermak; Prediction of extreme precipitation using a neural network: application to summer flood occurrence in Moravia. Adv. Eng. Software 31 (2000) 211 – 221. Broomhead, D. and D. Lowe; Multivariable functional interpolation and adaptive networks. Complex Syst. 2 (1988) 321–355. Cigizoglu, H. K.; Incorporation of ARMA models into flow forecasting by artificial neural networks, Environmetrics 14 (2003a) 417-427. Cigizoglu, H. K.; Estimation, forecasting and extrapolation of flow data by artificial neural networks. Hydrol. Sci. J. 48 (2003b) 349-361. Cigizoglu, H. K.; Estimation and forecasting of daily suspended sediment data by multi layer perceptrons. Adv. Water Resour. 27 (2004) 185-195. Cigizoglu, H.K., K. Alp and M. Kömürcü; Estimation of air pollution parameters using artificial neural networks, in: Advances in air pollution modelling for environmental security. NATO Advanced Research Workshop, 8-12 May 2004, Borovetz. Kluwer Academic Publisher, Dordrecht (2004). Cigizoglu, H.K. and O. Kisi; Flow prediction by two back propagation techniques using k-fold partitioning of neural network training data. Nordic Hydrol. 36 (2005) (in press).

Two Neural Network Methods in Estimation of Air Pollution Time Series

431

Duncan, K. and W.C. Potter; Ozone modeling using neural networks, J. Appl. Meteorol. 39 (2000) 291-296. Gardner, M. and S. Dorling; Artificial neural network – derived trends in daily maximum surface ozone concentrations, J. Air Waste Manage. Assoc. 51 (2001) 1202-1210. Hagan, M.T. and M.B. Menhaj; Training feedforward techniques with the Marquardt algorithm, IEEE Trans. Neural Networks 5 (1994) 989-993. Kuligowski, R.J. and A.P. Barros; Combined IR-microwave satellite retrieval of temperature and dewpoint profiles using artificial neural networks, J. Appl. Meteorol. 40 (2001) 205-2067. Luk, K.C., J.E. Ball and A. Sharma; A study of optimal model lag and spatial inputs to artificial neural network for rainfall forecasting, J. Hydrol. 227 (2000) 56-65. Maier, H.R. and G.C. Dandy; Neural network for the prediction and forecasting of water resources variables: a review of modeling issues and applications, Environ. Model. Software 15 (2000) 101-124. McKendry, I.G.; Evaluation of artificial neural networks for fine particulate pollution (PM10 and PM2.5) forecasting. J. Air Waste Manage. Assoc. 52 (2002) 1096-1101. Narasimhan, R., J. Keller, G. Subramaniam, E. Raasch, B. Croley, K. Duncan and W.C. Potter; Ozone modeling using neural networks, J. Appl. Meteorol., 39 (2000) 291-296. Sahai, A.K., M.K. Soman and V. Satyan; All India summer monsoon rainfall prediction using an artificial neural network, Clim. Dyn. 16 (2000) 291-302. Silverman, D. and J.A. Dracup; Artificial neural networks and long-range precipitation prediction in California, J. Appl. Meteorol. 39 (2000) 57-66. Seinfeld, J.H. and S. N. Pandis; Atmospheric chemistry and physics, John Wiley & Sons Inc., New York 1998, U.S. EPA (Environmental Protection Agency); Latest findings on national air quality: 2002 status and trends, http://www.epa.gov/airtrends, accessed on 15th February 2004. Wayne, R.P.; Chemistry of atmospheres, Oxford University Press, New York 2000.

List of Participants Aneja Viney P. Room 5136 Jordan Hall Campus Box 8208 Department of Marine, Earth, and Atmospheric Sciences North Carolina State University Raleigh, NC 27695-8208 USA

Phone: +1 (0)919 515-7808 Fax: +1 (0)919 515-7802 Email: [email protected]

USA

Anisimova Larisa Management Problems & Ecology Institute of Ukrainian Academy of Science DSAU Ukraine

Phone: +38 0567780483 Email: [email protected]

Ukraine

Argüello Gustavo A. INFIQC Dpto de Físico Química Fac. de Cs. Qcas., Universidad Nacional de Córdoba Argentina

Phone: + 54 (0)351 433 4169 / 4180 Fax:++ 54 (0)351 433 4188 Email: [email protected]

Argentina

Arsene Cecilia Faculty of Chemistry Dept. of Analytical Chemistry “Al. I. Cuza” University Iasi B-dul Carol I, No 11 Iasi –700506 Romania

Email: [email protected]

Romania

Barnes Ian Bergische University Wuppertal Fachbereich C –Physical Chemistry Gauss Strasse 20 42119 Wuppertal Germany

Phone: +49 (0)202 439 2510 Fax: +49 (0)202 439 2505 Email: [email protected]

Germany

Batchvarova Ekaterina National Institute of Meteorology and Hydrology 66, Tzarigradsko chaussee 1784 Sofia Bulgaria

Phone: +359 887507283 Fax: +359 2 9880380 Email: [email protected]

Bulgaria

Becker Karl H Bergische University Wuppertal Fachbereich C –Physical Chemistry Gauss Strasse 20 42119 Wuppertal Germany

Phone: +49 (0)202 439 3833 Fax: +49 (0)202 439 2505 Email: [email protected]

Germany

Brauers Theo ICG-II (Troposphäre) Forschungszentrum Jülich 52428 Jülich Germany

Phone: +49 (0)2461 616646 Fax: +49 (0)2461 615346 Email: [email protected]

Germany

Bronikowski Tadeusz

Phone: + 48 (0)22 3433223

Poland

433

434 Polish Academy of Sciences Institute Physical Chemistry Ul. Kasprzaka 44/52 Pl-01-224 Warszawa Poland

List of Participants

Fax: +48 (0)22 343 3448 Email: [email protected]

Carter William P. L. CE-CERT University of California Riverside 92521, California USA

Phone: +1 (0)951 781 5797 Fax: +1 (0)951 781 5790 Email: [email protected]

USA

Cigizoglu Kerem H. Istanbul Technical University Civil Engineering Faculty Division of Hydraulics Maslak 34469 Istanbul Turkey

Email: [email protected]

Turkey

Cocker David University of California, Riverside Bourns College of Engineering CE-CERT Riverside 92521, California USA

Phone: +1 (0)951 781 5695 Fax: +1 (0)951 781 5790 Email: [email protected]

USA

Crunaire Sabine Département Chimie et Environnement Ecole des Mines de Douai 941 rue Ch. Bourseul, BP 838 59508 Douai Cedex France

Phone: +33(0)3 27 71 26 12 Fax: Email: [email protected]

France

Dommen Josef Paul-Scherrer-Institute 5232 Villigen Switzerland

Phone: +41 (0)56 310 2992 Fax: +41 (0)56 310 2199 Email: [email protected]

Switzerland

Doussin Jean-Francois Maître de Conférence àl'Université Paris 7 Denis Diderot Equipe "Réactivité des Composés Traces en Atmosphère Réelle ou Simulée" Laboratoire Interuniversitaire des Systèmes Atmosphériques Faculté des Sciences 61, avenue du Général De Gaulle 94010 Créteil Cedex

Phone: +33 (0)1 45 17 15 85 Fax: +33 (0)1 45 17 15 64 Email: [email protected]

France

Gola Agnieszka Department of Physical Chemistry Medical University of Wroclaw Pl. Nankiera 1, 50-140 Wroclaw Poland

Phone: +48 (0)71 7840233 Fax: +48 (0)71 7840230 E-mail: [email protected]

Poland

Grüning Carsten European Commission Joint Research Centre (IES) TP 290, Via E. Fermi 1 I-21020 Ispra (VA) Italy

Phone: +39 0332 78 9767 Fax: +39 0332 78 9210 Email: [email protected]

Italy

Hjorth Jens Institute for Environment and Sustainability

Phone: +39 0332 789076 Fax: +39 0332 785837

Italy

List of Participants

435

(IES) The European Commission Joint Research Centre TP 290 I-21020 Ispra (VA) Italy

Email: [email protected]

Iinuma Yoshi Leibniz-Institut für Troposphärenforschung Permoserstr. 15 04318 Leipzig Germany

Phone: +49 (0)341 235-2988 Fax: +49 (0)341 235-2325 Email: [email protected]

Germany

Iwona Golonka Department of Physical Chemistry Medical University of Wroclaw pl. Nankiera 1, 50-140 Wroclaw Poland

Phone: + 48 (0)71 7840233 Fax: + 48 (0)71 7840230 Email: [email protected]

Poland

Jenkin, Mike Department of Environmental Science and Technology Imperial College Silwood Park Ascot Berkshire SL5 7PY UK

Phone: +44 (0)207 594 2514 Fax: +44 (0)207 594 2339 Email: [email protected]

UK

Kharytonov Mykola Management Problems & Ecology Institute of Ukrainian Academy of Science DSAU Ukraine

Email: [email protected]

Ukraine

Khodzher Tamara Limnological Institute of SB RAS Ulan-Batorskaya 3 Box 4299 664033 Irkutsk Russian Federation

Phone: +7 (0)3952 460 504 Fax: +7 (0)3952 460 405 Email: [email protected]

Russia

Kolev, Staytcho Scientific Secretary of the National Institute of Meteorology and Hydrology Bulgarian Academy of Sciences Blvd. Tzarigradsko Shaussee No. 66 1784 Sofia Bulgaria

Phone: +359 (0)2 751 018 Fax: +359 (0)2 988 0380 Email: [email protected]

Bulgaria

Le Bras Georges LCSR-CNRS 1C Av de la recherche scientifique 45071 Orléans la source France

Phone: +33 (0)2 38 25 5461 Fax: +33 (0)2 38 69 6004 Email: [email protected]

France

Lelevkin Valery Kyrgyz-Russian Slavic University 44 Kievskaya str. Bishkek, 720000, Kyrgyz Republic

Phone: 996 312 282909 Fax: 996 312 282776 Email: [email protected]

Kyrgyz Republic

Marinayte Irina. Limnological Institute of SB RAS

Email: [email protected]

Russian Federation

436

List of Participants

Ulan-Batorskaya 3 Box 4299 664033 Irkutsk Russian Federation McFiggans Gordon Atmospheric Physics Group, School of Earth and Atmospheric Sciences Manchester University M60 1QD, Manchester UK

Phone: +44 (0)161 200 3954 Fax: +44 (0)161 200 3951 Email: [email protected]

UK

Mellouki Wahid CNRS-LCSR 1c, avenue de la Recherche Scientifique F-45071 Orléans Cedex France

Phone: +33 (0)2 38 25 76 12 Fax: +33 (0)2 38 69 60 04 Email: [email protected]

France

Mihalopoulos Nikos Environmental Chemical Processes Laboratory Department of Chemistry University of Crete P.O. Box 1470, 71409 Heraklion Greece

Phone: +30 (0) 2810 393 662 Fax: +30 (0)2810 393 601 Email: [email protected]

Greece

Mocanu Raluca Faculty of Chemistry Dept. of Analytical Chemistry “Al. I. Cuza” University Iasi B-dul Carol I, No 11 Iasi –700506 Romania

Phone: +40232201354 Fax: +40232201313 Email: [email protected]

Romania

Monod Anne LCE, Case 29 Universite de Provence 3 Place Victor Hugo 13331 Marseille cedex 03 France

Phone: +33 491 10 6227 Fax: +33 491 10 63 77 Email: [email protected]

France

Niedojadlo Anita AGH University of Science and Technology Faculty of Mining Surveying and Environmental Engineering Department of Management and Protection of Environment Al. Mickiewicza 30, C4 30 059 Krakow Poland

Phone: Fax: +49 202 439 2505 Email: [email protected]

Poland

Olariu Romeo Faculty of Chemistry Dept. of Analytical Chemistry “Al. I. Cuza” University Iasi B-dul Carol I, No 11 Iasi –700506 Romania

Email: [email protected]

Romania

Phone: +48(0)22 343 3223 Fax: +48 (0)22 343 34 48 Email: [email protected]

Poland

Pasiuk-Bronikowska Wanda Polish Academy of Sciences Institute Physical Chemistry Ul. Kasprzaka 44/52

List of Participants

437

Pl-01-224 Warszawa Poland Paulo Pinho Departamento de Ambiente Escola Superior de tecnologia de Viseu Campus Politécnico 3500 Viseu Repeses Portugal

Phone: +351 (0) 232 480500 Fax: +351 (0)232 424651 Email: [email protected]

Portugal

Pilling Gwen Department of Chemistry University of Leeds Leeds LS2 9JT UK

Phone: +44 (0)113 2336 451 Fax: +44 (0)113 2336 565 Email: [email protected]

UK

Pilling Mike Department of Chemistry University of Leeds Leeds LS2 9JT UK

Phone: +44 (0)113 2336 451 Fax: +44 (0)113 2336 565 Email: [email protected]

UK

Rosenkranz Regina Bergische University Wuppertal Fachbereich C –Physical Chemistry Gauss Strasse 20 42119 Wuppertal Germany

Phone: +49 (0)202 439 2666 Fax: +49 (0)202 439 2505 Email: [email protected]

Germany

Rudzinski, Krzysztof J Institute of Physical Chemistry of the PAS Department of Catalysis on Metals Kasprzaka str. 44/52 01-224 Warsaw Poland

Phone: +48 (0)22 343 3402 Fax: +48 (0)22 343 3448 Email: [email protected]

Poland

Schady Arthur Wroclaw University of Technology Institute of Environment Protection Engineering I-15 (ZB-3) Wybrzeze S. Wyspianskiego 27 50-370 Wroclaw

Phone: +48 (0)71 320 2548 Fax: +48 (0)71 320 3599 Email: [email protected]

Poland

Sidebottom D University College Dublin Bielefeld, Dublin 4 Ireland

Phone: +353 (0)1 716 2293 Fax: +353 (0)1 706 2127 Email: [email protected]

Ireland

Sidebottom Howard University College Dublin Bielefeld, Dublin 4 Ireland

Phone: +353 (0)1 716 2293 Fax: +353 (0)1 706 2127 Email: [email protected]

Ireland

Siniaeva Tatiana Informational System of the State Ecological Inspection, Cosmonautilor, 09, MD 2005 Chisinau, Moldova

Phone: +373 (0)22 22 56 15 Fax: 373 (0)22 22 56 15 Email: [email protected]

Moldova

Smirnova, Tatyana Central Asian Research Hydrometeorological Institute 72, K. Maksumov str Tashkent, 700053

Email: [email protected]

Uzbekistan

438

List of Participants

Uzbekistan Solignac Géraldine LCSR-CNRS 1C Av de la recherche scientifique 45071 Orléans la source France

Phone: +33 (0)2 38 25 5058 Fax : +33 (0)2 38 69 60 04 Email : [email protected]

France

Tolkacheva, Galina Central Asian Research Hydrometeorological Institute 72, K. Maksumov str Tashkent, 700053 Uzbekistan

Phone: +988 (0)711331150 Fax: +988 (0)711331150 Email: [email protected]

Uzbekistan

Tomas Alexandre Département Chimie et Environnement Ecole des Mines de Douai 941 rue Ch. Bourseul, BP 838 59508 Douai Cedex France

Phone: +33 (0)3 27 71 2651 Email: [email protected]

France

Volkamer Rainer M Dreyfus Postdoctoral Fellow Massachusetts Institute of Technology; Room 54-1320 77 Massachusetts Avenue Cambridge, MA 02139

Phone: +1 (0)617-253-2321 Fax: +1 (0)617-258-6525 Email: [email protected]

USA

Wagner Robert Institut fuer Meteorologie und Klimaforschung, IMK-AAF POBox 3640 - Building 326 D-76021 Karlsruhe Germany

Phone: +49 (0)7247 82 5076 Fax: +49 (0)7247 82 4332 Email [email protected]

Germany

Wahner Andreas Forschungszentrum Jülich GmbH ICG-II: Troposphäre D-52425 Jülich Federal Republic of Germany

Phone:+49 2461 61 5932 Fax :+49 2461 61 4692 E-Mail:[email protected]:

Germany

Wenger John Department of Chemistry University College Cork College Road Cork/Ireland

Phone: +353 (0)21 490-2454 Fax: +353 (0)21 490-3014 Email: [email protected]

Ireland

Wiesen Peter Bergische University Wuppertal Fachbereich C –Physical Chemistry Gauss Strasse 20 42119 Wuppertal Germany

Phone: +49 (0)202 439 2515 Fax: +49 (0)202 439 2505 Email: [email protected]

Germany

Wirtz Klaus CEAM Parque Technológico Charles R. Darwin 14 46980 Paterna (Valencia) Spain

Phone: +34 (0)961318227 Fax: +34 (0)961318190 Email: [email protected]

Spain

List of Participants

439

Wojcik Dorota Department of Physical Chemistry Medical University of Wroclaw Pl. Nankiera 1, 50-140 Wroclaw Poland

Phone: +48 (0)71 7840233 Fax: +48 (0)71 7840230 Email: [email protected]

Poland

Zetzsch Cornelius Atmospheric Chemistry University of Bayreuth Dr.-Hans-Frisch-Str. 1-3 D-95448 Bayreuth Germany

Phone: +49 (0)921 555726 Fax: +49 (0)921 555729 Email: [email protected]

Germany

Zielinska Barbara Desert Research Institute Division of Atmospheric Sciences 2215 Raggio Parkway Reno, NV 89512-1095 USA

Phone: +1 (0)775/674-7066 Fax: +1 (0)775/674-7060 Email: [email protected]

USA

Author Index Albu Mihaela Alp Kadir Aneja Viney P. Anisimova Larisa B. Argüello Gustavo A. Arsene Cecilia

223 421 97 415 207, 213 121, 369

McDonald Jake McFiggans Gordon Mellouki A. Mihalopoulos Nikos Mocanu Raluca Möhler Ottmar Molina L. T. Molina M. J. Monod Anne

279 49 163 369 155, 223, 359 67 129 129 83

Babiy Andriy P. Barnes Ian Batchvarova Ekaterina Becker Karl Heinz Bejan Iustinian Bloss Claire Bloss William J. Blunden Jessica Bronikowski Tadeusz Bunz Helmut Burgos Paci Maximiliano A.

415 129, 155, 193, 223 301 1, 155, 341 155 143 143 97 253 67 207

Naumann Karl-Heinz Niedojadlo Anita Nishida S. Nolan Lorraine

67 341 223 171

Olariu Romeo-Iulian

121, 155, 369

Pasiuk-Bronikowska Wanda Pilling Michael J. Pinho Paulo G. Pio C. A. Platt Ulrich Poulain Laurent

253 143 241 241 129 83

Carter William P. L. Chen Boris B. Chiappero M. S. Cigizoglu Hikmet Kerem Claiborn Candis S. Cocker III David R. Coddeville Patrice Crunaire Sabine Cucu-Man Simona

27, 231 403 213 421 97 43 181 181 359

Rickard Andrew R. Rogers Hugo H. Rudzinski Krzysztof J.

143 97 261

Duncianu Marius

121, 369

Saathoff Harald Sagebiel John Schnaiter Martin Schurath Ulrich Seagrave JeanClare Sidebottom Howard Smirnova Tatyana Song Chen Spassova Tatiana Steinnes Eiliv Stockwell William

67 279 67 67 279 171 393 43 301 359 279

Fittschen Christa

181

Grigorieva Vera Gritsan Natalia P. Guihur Anne-Laure

351 415 171

Hurley M. D.

213

Iordanova Liliana

301

Jenkin Michael E.

143, 241

Takahashi K. Tolkacheva Galina Tomas Alexandre

213 379, 393 181

Kharytonov Mykola M. Kolev Staytcho Kömürcü Mügé Kurtenbach Ralf

415 351 421 341

Valkov Nedialko Volkamer Rainer

301 129

Lelevkin Valery M. Lemoine Bernard Linke Claudia

403 181 67

Malanca F. E. Manning Marcus Matsumi Y.

213 171 213

Wagner Robert Wallington Timothy J. Wenger John C. Wiesen Peter Wirtz Klaus Wortham Henri

67 213 111 279, 285, 295, 341 121, 279 83

Zielinska Barbara

279

441

Subject Index Absorbent tube Absorption – long path Absorption cross section - UV Acetaldehyde (CH3CHO) Acetals Acetate Acetic acid Acetone Acetonitrile Acetylene Acidity precipitation/pH/water Acrolein Actinic flux Adduct formation Adiabatic expansion Aerosol – atmospheric/general/transport – chemical composition – condensation processes – extinction spectra – Junge layer – mass concentration – mass spectrometer – physical properties – proxies – secondary organic (SOA) – size distributions – stratospheric – studies Agricultural soils AIDA aerosol and cloud chamber Air glow – reaction Air quality – monitoring/pollution Airshed model Air-water interface Alcohols Aldehydes Aldehydes Alkanes Alkenes Alkoxy radical Alkyl halides Alkyl iodides Alkyl nitrates Alkylbenzene Alkynes

342 1 111, 130, 137, 219 8, 113, 123, 126, 242, 246 264 134 181 348 86 135 98, 369, 374, 398 113, 126, 147 112, 121, 137 136 69 49, 301, 307, 310, 311, 331, 370, 385, 387, 393, 400, 403 49, 256 307 85 406 46 57, 69 49, 61, 67, 301, 307, 404 61 27, 31, 40, 43, 69, 129, 143, 151, 155, 163, 268, 298, 301, 303, 324, 393 31, 50, 56, 307, 310, 405 405 3, 263, 301 98 3, 67 13 27, 98, 301, 303, 306, 309, 311, 312, 314, 318, 323, 327, 329, 359, 369, 376, 379, 392, 398, 408, 415, 421 27, 231 98 341, 346 37, 111, 287, 293 8, 113, 114, 123, 126, 144, 242, 246, 264 21, 135, 200, 287, 341, 346 37, 18, 37, 134, 164, 171, 200, 225, 341, 346 171 193, 324 193, 324 214 135 346 443

444

mmonia (NH3) Ammonium sulphate (NH4SO4) ANN methods Anthropogenic emissions Aqueous-phase transformation Aromatic aldehydes Aromatic hydrocarbons Aromatic hydrocarbons – ring retaining/opening products Arrhenius plot – rate constants Arsenic Atmosphere – chemistry studies/marine Atmosphere – biosphere interactions Atmospheric – pollution

Subject Index

57, 97, 103, 311, 315, 320, 323 51, 70, 263 421, 423 98, 155, 181, 261 83, 253, 264, 272 114, 144 37, 44, 134, 143,144, 155, 181, 231, 236, 254, 286, 324, 330, 341, 346 133, 144, 148 85, 164 327, 359, 394 1, 129, 223, 295, 301, 310, 316 97, 272

– brown cloud – lifetime/NO3/OH – reactivity Autoxidation

27, 98, 301, 303, 306, 309, 311, 312, 314, 318, 323, 327, 329, 359, 360, 363, 369, 376, 379, 392, 398, 408, 415, 421 412 118, 130 231 253, 256, 266

Backscattering coefficient/ratio Backtrajectory Benzaldehyde Benzene Benzoquinone Bicyclic species Bicycloalkyl radical Bicycloalkylperoxi radical Bio-accumulation Biodiesel Biogenic emissions/sources Biomass burning Bio-monitoring Bio-solids Branching ratio Brazil – Porto Allegre Bromine oxide (BrO) Buffer capacity Bulgaria Butane (n-) Butanoic acid Butanol (2-) Butene (1-) Butene (iso-) Butene (trans-2-) Butenedial

403, 406, 408 403, 405, 410, 411 144 130, 133, 143, 155, 327, 347 144 151 133 133 359, 363 286, 290 98, 163, 181, 261, 381 261 359, 363 98 181, 189, 213, 217, 225, 228 133 57 415 301, 303, 351 19, 32, 156, 241, 249, 286, 289 90 348 347 225, 347 37, 135 145

Subject Index

445

Butyl acetate Butyl vinyl ether Butyl vinyl ether (iso-) Butyl vinyl ether (tert-) Butyraldehyde

348 163 163 163 113, 126

Cadmium Calcium Car exhaust Carbon bond mechanism Carbon dioxide (CO2) Carbon monoxide (CO)

320, 327, 359, 363 311 279, 285 235 53, 59, 271, 285, 375, 379, 422 2, 32, 53, 57, 147, 231, 237, 262, 287, 302, 327, 397, 417, 421, 429 121, 134, 181, 208, 242, 262, 292, 341 207 343 342 147, 181, 373, 388 261 126 132 256, 266 181

Carbonyl compounds Carbonyl fluoride (CF2O) Carbosieve Carbotrap Carboxylic acids Carcinogenic Caronaldehyde Cartridge collection Catalyst Cavity ring down spectroscopy (CRDS) CCD camera Central Asia CF3 CF3N2NO2 CF3O CF3ONO2 CF3OO CF3OONO2 CFC Chamber – aerosol – AIDA – characterisation – dynamic flow through – effects – EUPHORE – indoor – outdoor – radical source – SAPHIR – studies Chemical activating/deactivating effects Chemical composition Chemical mass balance modelling

80 403, 409 214 215 213 213, 215 213, 215 215 207 49 67 32, 37 97 31, 242, 258 4, 22, 50, 112, 121, 143, 171, 241, 279, 285, 324 1, 2, 3, 43, 50, 55, 183, 207, 225, 231, 242 2, 4, 27, 34, 50, 111, 171, 279, 285 33 5, 34 111, 207, 231, 241, 324 160 394 341, 345, 348

446

Subject Index

Chemical mechanism – SAPRC-99 Chemical mechanisms Chemical modelling Chemical processes Chloral (CCl3C(O)H) Chloride Chlorine (Cl2) Chlorine atom trap Chlorine nitrate (ClONO2) Chloroalkoxy radicals Chloroglyoxal Chromium Cirrus cloud Cl atom Cl2Climate Cloud chemistry Cloud condensation nuclei (CCN) Cloud droplets Cloud processes Coagulation Coatings VOCs Combustion processes Condensation Condensation particle counter Continuously stirred tank reactor (CSTR) Copper CRDS – operation principle Cresols Criegee radical Crops Cross-activation Crotonaldehyde Czech Republic

32, 37, 50 27, 83, 94, 231, 386 4 1, 49 171 311 21, 52, 172, 253, 320 173 52, 253, 256 171 173 419 67 7, 20, 52, 83, 172 83 129, 223, 271, 301, 361, 370, 379, 393, 403 67, 331 58, 273 49, 83, 163, 256 54 55, 68 231 155 55 57 99

Databank Deliquescence Deposition – wet and dry

309 61 139, 181, 311, 316, 365, 369, 371, 376, 379, 385, 393 132 29 144, 148 280 279, 282, 285 280 56, 72 22, 57, 129, 144, 313

Derivatisation agent Dew point Dicarbonyls Diesel engine Diesel exhaust Diesel fuel Differential Mobility Particle Sizer Differential optical absorption spectrometer (DOAS)

360, 418 186 144, 148, 155 265 97 253 113, 126 301, 306

Subject Index

Diffusion – aqueous Diffusion – gas Diffusion reflectance infrared Fourier transform (DRIFTS) Dihydroxybenzenes Diiodomethane (CH2I2) Dimethyl sulphide (CH3SCH3) Dimethylbenzaldehydes Dinitrates Dinitrophenylhydrazine (DNPH, 2,4-) Diode laser Dobson spectrometer/unit Double salts Droplets Dust Ecosystems Efflorescence Electron spray ionisation Electrophilic addition Elemental carbon Emission – industrial/profile/sources/trends – database – chlorinated solvents – heavy metals – methanol – swine waste – vehicles Energy transfer Engine test rig Environmental Chamber Environmental chemistry – new EU member states EPA Epoxy compounds Esters Estonia Ethane (C2H6) Ethene (CH2=CH2) Ethers Ethyl vinyl ether EU member states – new EUPHORE chamber EUROCHAMP – project Europe

447

51 51 263 155 50, 62 62, 223 114 262 132 181 325, 351, 407 61 49, 83, 163, 253 381, 3957, 403, 408, 421 171, 181, 213, 241, 261, 301, 341, 351, 359, 369, 379, 393, 403, 415, 421 61 88 160 55 341, 346, 349, 369, 379, 382, 392, 499, 415, 419, 422, 426 307, 341 171 363 207 97 129, 143, 279, 285, 309, 389 10 291 1, 2, 3, 4, 27, 34, 43, 50, 55, 111, 171, 183, 207, 225, 231, 241, 253, 258, 264, 266, 279, 285, 295 301 4, 342 145 134, 341, 346 301, 309 286 18, 37, 156, 193, 225, 286, 289 134, 163, 341 163 301 4, 22, 50, 112, 121, 143, 171, 241, 279 24, 295 1, 295, 341

448

Subject Index

European Environment Agency ((EEA) Evaporation Expansion cooling Extinction Eye irritation

301 55 78 403 1

Feed forward back propagation (FFBP) Feedback Fenton reaction FEP foil Fertilizers Field campaign – MCMA-2003 Field measurement – BERLIOZ Field studies Flow tube Fluorinated compounds Fluorinated peroxy radicals Fluorine (F2) Flux – calculation/general Flux – gas Flux – models Fog droplets Formaldehyde (HCHO) – photolysis and offgasing Formamide Formates – organic Formyl-2-pyrrolidone (FP) Fragmentation pathway Free radicals FTIR spectroscopy Fuel additives Fuels Furanone

411, 423 261, 270 84 3, 4, 28, 29, 54, 97, 99, 172 97 130 7 301, 345 13 207, 293, 416 213 320 100, 371, 378 98 104 163, 331 8, 30, 32, 35, 37, 57, 144, 165, 242, 245, 285, 293, 315, 320 90 165, 167 89 89 83 1, 15, 22, 73, 112, 144, 156, 172, 181, 208 163, 194, 207, 215, 280 181, 280, 285 144

Gas analyser Gas chromatograph (GC) - general

57 15, 30, 112, 132, 144, 172, 181, 194, 197, 264, 280, 343 44, 50 3 308 101 97, 254, 262 285 185 133 3 314

Gas particle partitioning Gas phase – studies Gas phase ion chemistry Gas uptake – plants Gas-liquid interface/transfer Gasoline GC-FTIR Germany – Pabsthum Glass – Duran/Pyrex/quartz Global Ozone Monitoring Experiment (GOME) Glyceraldehydes-3-phosphate

261

Subject Index

449

Glycerine Glycoaldehyde Glycol derivates Glyoxal Green house gas Grosse Bonner Kugel

208 113, 126 341 113, 126, 129, 133 396, 403 1, 10

H atom H2SO4/H2O solution droplets Halocarbon Halocarbon solvents Halogen cycling Halogenated compounds Halogens HCFC HDO determination Health – effects/human

20 69, 75 208 163 52 214, 341 293 207, 213 181 1, 43, 129, 155, 163, 261, 279, 301, 351, 359, 363, 369, 293, 415, 419, 421 303, 307, 313, 317, 327, 359, 363, 368, 379, 394, 401, 415, 418 264 269 1 55, 67, 207, 242, 253, 261 113 145 207, 213 207 85

Heavy metals Hemiacetals Henry’s law Herriott cell Heterogeneous processes Hexanal Hexene-2,5-dione (3-) HFC HFE High pressure liquid chromatography (HPLC) HNO3/H2O solution droplets Homogeneous processes Humic-like substance (HULIS) Humidity measurements Hungary Hydoxyl radical (OH) Hydoxyl radical (OH) – dark source Hydoxyl radical (OH) – measurement Hydoxyl radical (OH) – photolytic source Hydrocarbon/NOx ratio Hydrocarbonates Hydrocarbons Hydrochloric acid (HCl) Hydrogen fluoride (HF) Hydrogen peroxide (H2O2) Hydrogen sulphide (H2S) Hydroperoxy radical (HO2)

75 55 263 72 301 1, 4, 19, 22, 83, 111, 130, 143, 155, 163, 172, 181, 193, 207, 223, 262, 279, 396 7,18 24 7, 9 43, 46 388 43, 61, 163, 181, 341, 383, 422 171, 178, 320 315 30, 97, 104, 172, 226, 264, 395 97, 303, 315, 320, 327 2, 4, 10, 22, 143, 262, 397

450

Subject Index

Hydroscopic growth factor Hydroscopicity – particles Hydroscopicity Tandem Differential Mobility Analyser (HTDMA) Hydroxyl aromatics Hydroxymethylpyrrolidone (NHMP) Hydroxy-N-methylpyrrolidone (5-HNMP) Hypoiodous acid (HOI)

51 53 57

Iasi Ice crystals Ice nucleation Incremental reactivity – maximum (MIR) Index Atmospheric Pollution (IAP) Industrial waster utilization Infrared (IR) absorption Infrared spectroscopy Infrastructure Inorganic salts Integrating activities iodine Iodine dioxide (OIO) Iodine monoxide (IO) Iodoalkanes Iodopropane Ionisation potential Ions Iron ore Isododecane Isomerization Isopentene Isoprene Isopropanol

369 49, 70 67, 70, 76, 77 4, 231, 293

295 55 295 62, 193 57, 62 52, 62, 194 194, 200 195 21 253, 261, 317, 331, 372, 376, 387, 388, 395 416 226 136 347 50, 103, 129, 135, 156, 261, 268, 286 173

Joint research activities

298

K2S2O8 Kelvin effect Kinetic – studies

266 60 1, 3, 15, 19, 83, 157, 165, 167, 169, 193, 200, 223, 227, 231, 289

Lactone – D-angelica Lagoon Lamber Beer equation Lamp – argon arc Lamp – black Lamp – fluorescent

145 99 73 29, 31, 44, 59 15, 28, 29, 31, 44, 59, 172, 215 15, 172, 181, 196, 197, 215, 226

155 90 90 201

381, 390 304 16, 209, 217

Subject Index

451

Lamp – UV Lamp – xenon arc Latvia Lead Letovicite Lidar Lifetime Light flux Liquid hog waste Lithuania Los Angeles (LA)

226 58 301 315, 320, 327, 359, 363, 394 71 403 211, 213, 215, 220, 279 59 98 301, 314 1, 133

Macroalgae laminaria Macromolecular substances Macromolecule Magnesium ore Maleic anhydride Mass spectrometric detection Master Chemical Mechanism (MCM) Maximum admissible concentration (MAC) Mechanism – aqueous phase – evaluation – gas phase

62 263 53 416 146 83, 88, 132, 310 131, 135, 143, 231, 241, 262 390

– RACM – SAPRC-99 Mechanistic – studies Mercury Metallurgical industry Metals Methacrolein Methane (CH4) Methanol Methyl bromide (CH3Br) Methyl chloride (CH3Cl) Methyl ethyl ketone (CH3COC2H5) Methyl nitrite (CH3ONO) Methyl tert-butyl ether (MTBE) Methyl vinyl ether Methyl vinyl ketone Methyl-2-nitrophenol (3-) Methyl-2-nitrophenol (4-) Methyl-2-nitrophenol (5-) Methyl-2-nitrophenol (6-) Methyl-4-aminobutanoic acid Methylamine Methyl-butyraldehyde

94, 255, 257, 267 27,37, 143, 148, 231, 236 3, 27, 143, 174, 175, 176, 177, 178, 193, 200, 217, 224, 231, 236, 241, 244, 262, 286 286 37, 231, 242 3, 143, 193, 200, 217, 397 317, 359, 363 416 303, 307, 313, 317, 327, 359, 363, 368, 379 113,126, 135, 262 97, 286, 417, 422 207, 263 422 422 84, 242, 247 156, 193 342, 346 163 126, 129, 262 159 159 159 157 90 90 113, 126

452

Methylglyoxal Mexico City Microbiological properties Microphysical processes Mie theory Mineral dust Minerals MIR – maximum incremental reactivity Mirror system MnSO4 Model – photochemical Modelling – coupled Modelling – pollution MOIR/2 Monitoring- networks/sites Moss types/monitoring MPAN Multi-phase processes Mutagenic compound N atom Na2SO3 Nano particle Networking activities Nickel Nitrate anions Nitrate radical (NO3) Nitrates Nitric acid (HNO3) Nitric acid dehydrate (NAD) Nitric oxide (NO) Nitric oxide (NO) - reaction Nitroaromatic compounds Nitroaromatics Nitrocresols Nitrogen Nitrogen dioxide – dimer (N2O4) Nitrogen dioxide (NO2) Nitrogen dioxide (NO2) – photolysis Nitrogen pentoxide (N2O5) Nitrophenols Nitrous acid (HONO) Nitrous oxide (N2O) Nitroxyalkyl (peroxy radicals) N-methyl-pyrrolidone (NMP)

Subject Index

133 130, 133 98 49 57 80 55, 387, 393 233 1 266 138, 241 49, 54 305, 309, 311, 314, 317, 323, 326, 328, 377, 388, 422 233 301, 303, 306, 309, 312, 314, 318, 323, 327, 329, 332, 361, 363, 371, 380, 392, 403, 415, 421 359, 364 262 67, 69, 261, 268 129 10 267 57 297 327, 359 254, 395 1, 47, 53, 57, 83, 111, 155, 163, 169, 253, 262, 279 163, 262, 311, 388, 398, 401 30, 32, 144, 253, 289, 298, 311, 373 70, 74 19, 21, 30, 32, 57, 97, 103, 129, 156, 213, 236, 280, 286, 306, 315, 369, 381, 394, 401, 411, 427 10,13 155 155 155 97, 311, 395 219 20, 57, 69, 97, 129, 144, 253, 280, 303, 306, 313, 315, 318, 323, 381, 391, 394, 401 8, 30 40, 69 155 8, 32, 34, 151, 235, 288, 293 422 262 83

Subject Index

N-methylsuccinimide (NMS) NO/NO2 ratio Nonanal Non-methane hydrocarbons (NMHC) Norrish type I mechanism Norrish type II mechanism NOx NOx - offgasing NOx-limited NOy Nuclear explosion O atom Octane (n-) Olefin Organic nitrate Oxidation capacity Oxo-2-pentenal (4-) Oxo-but-2-eneperoxoic acid (4-) Oxo-but-2-enoic acid (4-) Oxygen Oxygenated compounds Ozone – layer –surface – total/stratospheric – troposphere/photochemical

– creation potentials – pollution Sofia Ozonides Ozonolysis Particle physics Particulate matter (PM) Pentafluorobenzyl hydroxylamine (PFBHA, O-2,3,4,5,6-) Pentan-2-one Pentanal (n-) Perfluoroperoxyacetyl nitrate Peroxy acetyl nitrate (PAN) Peroxy acetyl radical (CH3CO3) Peroxy nitrate (RO2NO2) Peroxy radical (RO2) – reaction/source Peroxynitric acid (HO2NO2) Peroxynitric acid (HO2NO2)

453

86, 89 286 113 29, 57, 155, 341, 348 114 114 3,5, 44, 45, 133, 231, 233, 236, 241, 265, 280, 286, 311, 369, 381, 393, 417, 422 27, 31, 33, 35 233 30, 97 409 10, 13, 193, 214, 324, 396 37 10, 167 213 149, 207, 393 145 147 147 8, 10, 216, 278, 386, 398, 405 135, 152, 163, 181, 341, 346 8, 351 351, 354, 385 2, 351, 354, 405, 407 1, 27, 28, 30, 32, 33, 53, 69, 111, 129, 143, 163, 167, 214, 231, 236, 241, 261, 279, 287, 301, 306, 311, 313, 315, 318, 327, 351, 354, 356, 379, 396, 398 143, 293 353 262 68, 262, 324 310 28, 43, 53, 279, 282, 285, 288, 303, 306, 310, 315, 318, 326, 327, 385, 421, 427 132 126 113, 126 30, 249, 272, 279, 289, 320, 395 181 18, 20, 22 7, 21, 214, 389 19, 21 19

454

Pesticides Phenols Phosgene (Cl2CO) phosphorus Photocatalysis Photochemical model Photochemistry

Subject Index

Products OH + CH3OOD Propene (CH3CH2=CH2) Propionaldehyde Propyl iodide Propyl vinyl ether Pyrrolidone (2-) Pyruvate Pyuvic acid

330, 362 145, 148, 155, 315, 330 171 311 86 131 1, 54, 58, 83, 130, 143, 155, 163, 193, 213, 218, 288, 351, 387, 393, 397 83 1, 111, 121, 129, 155, 171, 193, 213, 287, 396 111, 121, 126, 127, 136, 158, 243, 287 113, 118, 171 146 111 54 3 111, 155, 237, 261 1, 2, 3, 15, 155, 194, 264, 279, 285 17 1, 129, 147 1 261 389, 395 68, 241 136 113, 126, 253 113, 126 1 1, 98, 261, 271 61 301, 318 67 283, 320, 422 283 133 311 98, 311, 369, 374, 379, 385, 388, 394, 399, 401 111, 129, 143, 163, 173, 193, 210, 213, 216, 241, 262, 288 181, 189 37, 134, 225 113 193 163 86, 90 261 126

Quantum yield

112, 113, 127, 136, 171, 215

Photochemistry – aqueous phase Photolysis Photolysis frequency Photolysis lifetime Photolysis mechanism - butenedial Photolysis rate coefficient Photon flux Photooxidant - studies Photooxidation Photoreactor Photoreactor – themostated Photosmog Photosmog formation Phytoplankton Phytotoxic compounds Pinene (D-) Pinene (E-) Pinonaldehyde Pivaldehyde Plant - damage Plant(s) Plume Poland Polar stratospheric cloud Polyaromatic hydrocarbons (PAH) Polymeric substances Portugal – Giesta Potassium Precipitation Product yields/formation

Subject Index

Radial basis function Radiative transfer Radical amplifier Radical chain Radical source Radical(s) Rape oil methyl ester Rate constant – C2H4 + OH Rate constant – DMS + OH Rate constant – HO2NO2 + OH Rate constants – determination Rate constants – nitrophenol Reaction products Reactors Receptor points Reduced organic sulphur compounds Regional atmospheric chemistry mechanism (RACM) Relative humidity Relative rate technique Reservoir species ROG – Reactive organic gas ROG surrogate ROG/NOx ratio Romania Salt particles SAPHIR chamber Satellite observations Scandium ore Scanning electrical mobility spectrometer Seasonal trends – pollutants Sediment Seed aerosol Semi-volatile compounds Sensitivity analysis Silver fluoride (AgF2) Simulation Simulation – model Simulation – plant damage Slovak Republic Slovenia Smog chamber SO4Sodium chloride (NaCl)

455

421, 425 54 237 1, 9 1 1, 20, 261, 298 286 15 227 15, 19 1, 15, 19, 83, 157, 165, 167, 169, 193, 200, 223, 227, 231, 269 158 1, 83 1, 2, 3, 4, 27, 34, 43, 50, 55, 111, 171, 183, 207, 225, 231, 241, 253, 258, 264, 266, 279, 285, 295 349 97 286, 289 51, 55, 59, 78 9, 84, 157, 193, 197, 223 254, 366 37, 231, 233 37, 39 38, 231 301, 323, 361, 371 51, 62 5, 34, 50 69 416 45, 57, 282 383 362 78 51, 283 145, 153, 236, 291 215 1, 286 32, 146 1 301, 327 301, 329 4, 241 83, 254, 395 51, 62, 253

456

Sodium nitrate Sodium sulphite Soils Solar eclipse Solar flux/radiation Solar spectrum Solar zenith angle Solid absorbent Solvent – chlorinated Solvent applications/profiles Solvent extraction Soot particles Source apportionment/general Spectroradiometer Spectroscopy IR Stirred tank reactor Stratosphere Structure activity relationship (SAR) Succinimide Sulphate Sulphide Sulphite ion Sulphite radical Sulphoxy radicals Sulphur Sulphur compounds Sulphur dioxide (SO2) Sulphuric acid (H2SO4) Sunlight Supercooled ternary solution (STS) Supersaturation Surface reaction Surface tension Surrogate – mixture Swine waste Tandem differential mobility analyzer Tashkent City/Province Teflon foil Temperature control - chambers Terpenes Terpenoids Tetrachloroethene Tetraols Thermal decomposition Thermal decomposition – peroxynitrates

Subject Index

253 266 361, 393, 415, 419 357 112, 370, 386, 393 30, 55 122, 130, 136 132 171 143, 163, 181, 341, 348 132 68, 71, 78, 285, 408 341, 393 59, 121, 287 1 266 351, 403 4, 165 86 68, 253, 261, 267, 311, 315, 369, 386, 394, 398, 401 97, 303, 315, 320, 327, 387 267 267 253 97, 254, 286 54, 393 57, 62, 254, 266, 285, 303, 311, 313, 315, 318, 323, 327, 369, 373, 379, 386, 390, 394, 405, 407, 417, 422 69, 75, 254, 263, 265, 373 111, 129 69 54, 58 208 61 37, 232, 286 98 45, 52 379, 393 1,15, 121, 181, 193, 263, 273, 279, 352 3, 15, 59, 67 194, 272 272 171 265 213 18

Subject Index

Thermal desorption system Tolualdehydes Toluene Toxicity Traffic – road Traffic – tunnel Trajectories – air mass Trichloroethane (1,1,1-) Trichloroethene Trimethylbenzene (1,3,5-) Troposphere Tropospheric lifetimes Tunable diode laser absorption spectroscopy

457

184 114 37, 44, 130, 133, 135, 143, 148, 155, 236, 242, 286, 289, 347 279, 283, 359, 379 279, 285, 341, 345, 352, 387, 395 345 373, 403, 410 171 171 50, 143, 148, 242 2, 8, 143, 155, 163, 218, 261, 351, 403 169, 171 30, 57, 73

UCR EPA environmental chamber Ukraine Uranium ore Urban air USA UV absorption spectroscopy Uzbekistan

27 415 416 132, 379, 393, 415 133, 281, 296 15, 73, 213, 265, 357 379

Vegetation Vehicle exhaust/emissions Vinyl esters Vinyl ethers Visible absorption spectroscopy VOC/NOx ratio Volatile organic compounds (VOCs)

97, 261, 272, 352, 389, 393, 415, 421 129, 143, 181, 261, 279, 285, 318, 395 166 163 15, 73 4, 148, 285 4, 27, 33, 111, 121, 129, 148, 155, 163, 181, 208, 231, 241, 271, 279, 285, 298, 320, 417 405, 421

Volcanic eruption Water affinity Water soluble organic compounds (WSOCs) WATER9 model White cell optics Water – determination/ rain/river/surface Wastewater

56 83 105 1, 15, 58, 73, 156, 172, 194, 226 189, 359, 365, 370, 373, 379, 396, 416 362

Xylene (m-) Xylene (p-)

37, 43, 236, 347 130, 135, 143, 148, 242, 347

Zinc

360